14:36
Please welcome Ivan Jiao.
14:50
Hi again. Good morning and welcome to Make Notion 2025.
14:59
Thank you for joining us here in San Francisco while watching online around the world.
15:07
You know, last year's Make With Notion was our very first conference as a company
15:13
and to be honest, we're not exactly sure how it would turned out. So, raise your
15:18
hand if you were there last year.
15:23
So, I think you can confirm that it turned out beautifully.
15:29
Many people told me that it felt like a coming of age party for notion and for our community.
15:36
I was especially proud of the creativity and the care that our team put into the
15:41
event. You know, at Ocean Notion, we love well-crafted things. We're a little
15:47
bit snobby, especially about chairs. So to make things a little bit more
15:53
interesting last year, we brought our nicest chair from our office to the event.
16:00
And some people joke that it felt more like an interior design conference than a tech conference and I'll take that as
16:07
a compliment. So for today, I hope you can have a
16:14
similar great experience as last year, if not better. We prepared a lot of exciting programs
16:21
and demos for you. I think you'll have a blast.
16:27
So, let's kick off today's keynote with a product recap. We launched a lot of
16:32
things last year, including brand new AI products. But I want to start with something
16:38
that's not as flashy, but critical to all of you. Performance and reliability.
16:46
We really care about this as Notion continue to scale. So our engineer team
16:51
spent the entire summer on this. We fixed over 2,000 bugs.
16:58
We improved desktop speed by 30%. And for large databases, they can now
17:04
load up to three times as fast as before. I think you can all feel the
17:10
difference. Thank you.
17:17
For our developer community, we ship the top API requests for years, web hooks.
17:24
So you can now build more powerful integration around notion for your enterprise, for your products.
17:32
Also, we launched something that many of you have been waiting for ever since we
17:38
started this company.
17:44
Offline mode. So, last year on stage, I teased offline mode to you guys
17:52
and last month we ship offline mode to all of you.
18:01
Well, since then, our subreddit traffic has been plummeting.
18:07
Uh, some Reddit users are wondering what to talk about now. Notion offline mode
18:13
is out. But to be clear, there's a lot more we
18:18
need to do to make Notion and Notion offline even better. And thank you again
18:23
for your patience and for your feedback. This is the reason we love our community
18:29
so much because you all care so much.
18:34
We also launched brand new AI products like AI meeting notes so you can Thank
18:40
you. My favorite too. Yeah. Um which does the note takingaking so you can
18:45
focus on being present in the meeting. And there's enterprise search. It turns
18:51
your company's tribal knowledge into instant answers.
18:56
We also launched notion mail on iOS and and iOS and desktop. Android is coming
19:02
soon. And last but not least, we upgrade notion calendar with scheduling link and
19:10
task integrations. Yeah. And in today's keynote, you'll learn
19:16
that how all these products tools come together with notion and with AI.
19:25
Last year was also a big year for our customers with more companies and teams
19:31
adopting notion like cursor. It's one of the fastest
19:36
growing company of all time and it runs on notion
19:41
or openai it also runs on notion. They on board thousands of new hire last year
19:48
and when those new hireer have questions they ask notion AI. It's pretty neat to
19:54
see that the company that start the current AI revolution rely on notion AI.
20:01
And when you look at this new generation of AI companies I'm really proud that
20:07
almost all of them star on notion.
20:14
So outside AI industry, RAMP is a fast growing company reinventing finance. And
20:21
I especially like their tagline, time is money. Save both.
20:27
Last year, Notion helped RAMP consolidate half a dozen tools and save
20:32
70% of their tooling cost. And this is my favorite. To quote our
20:38
CEO, Eric, Notion is not just saving them money, but literally compressing
20:44
time. So now I think about it. Well, maybe notion should steal RAM tech line.
20:54
Of course, we'll never forget small business and creators.
20:59
Take pies and thigh, a restaurant in Brooklyn. And according to the internet,
21:05
they have the best pies in New York City. And I finally got to taste it this week. It's excellent.
21:12
They're scaling and opening new locations. And they're using Notion for almost everything from training to
21:19
scheduling to brainstorming new recipes. So next time when you have their pies or
21:26
thighs, maybe there's a little bit AI in it.
21:33
Notion is also growing in the education and this one is very personal for me.
21:40
Some of you probably know notion was reborn in Kyoto, Japan. Were inspired by
21:46
Kyoto's arts and craft. And today Kyoto University of Arts uses
21:53
notion across 800 faculty. It's very heartwarming to see that the
21:59
school that teaches Kyoto's and arts and craft now runs on notion.
22:06
Thank you.
22:12
All the customer story comes back to our mission. Whether you're building an AI
22:17
startup, reinventing modern finance, or expanding your restaurant operations, we
22:24
want notion to be there. to support you with beautiful tools to
22:30
help you build your life's work. Notion the company is also growing.
22:37
Last year, we opened five new locations. Sinei, Singapore, London, Paris, and
22:44
Munich, so we can stay closer to our community and customers.
22:50
In San Francisco, we also move into a new office.
22:55
It's one of the first office building in the city and it survived 1906 San
23:01
Francisco earthquake. So if you zoom in to uh our building on
23:06
the right, it says offices ready in May. So right
23:13
after the San Francisco earthquake, people were already eager getting back to office.
23:20
So return to office is different back then. So, so time has clearly changed,
23:26
right? So, you might be wondering, well, why am
23:32
I talking about our office? Because it relate to the theme of our
23:38
keynote today. Around 1900,
23:44
most Americans were farmers, but people began to migrate into cities for a new
23:50
type of work, knowledge work.
23:56
Office building were created to host knowledge workers and new tools were
24:02
designed to automate people's workflows. tools like typewriters so people don't
24:08
didn't have to second guessess their co-workers handwritings or file cabinets which was basically the
24:16
original second brain and now imagine creating entire YouTube course about file cabinets
24:26
as we move into the software era typewriters become word processors
24:32
file cabinets became databases and applications And today the software era became the
24:40
software spraw era tools are fragmented and rigid.
24:48
Average company need to use ADA tools to run their business and knowledge work feels more like busy
24:56
work. So sometimes we think well busy work
25:01
it's just modern life you know part of a being adult grow up
25:08
but at notion our mission is to help you do your life's work so we ask ourself
25:16
can we use AI to help you delegate some of your busy work
25:23
and as you probably know AI and AI agent already exists for some sectors like
25:29
customer support and coding, but the world has yet to see a AI agent
25:35
for general purpose knowledge work. There are a lot of cool demos on Twitter, but nothing truly works at
25:41
scale reliably. This is the problem we've been working
25:47
on for the past two years. And sometimes it feels everything we
25:52
built at Notion has prepared us for this moment.
25:57
So what does it take to build a knowledge work agent? It needs three parts.
26:03
First, it needs to know about your work. No, context is everything for AI.
26:10
But today, here's the dilemma of AI at work. On one side, AI chatbot knows
26:15
about your personal life, but it's not plugged into your work context. And on the work side, every SAS app has
26:23
AI now, but they don't work together. So the context is still fragmented.
26:30
At Notion, we've been solving this fragmentation problem for years.
26:36
With Notion 1.0, we consolidate documents and knowledge base into one
26:41
tool. And with 2.0, we're introducing databases and brought together tasks and
26:47
project and your other workflows. Then we have notion mail, notion
26:54
calendar and AI meeting notes. For everything else there is notion
27:00
enterprise search. This means notion knows everything about
27:06
you and your work and we can pass this contact to your notion knowledge work
27:11
agent. Second, to build a knowledge work agent,
27:18
it need to be designed for teamwork. But most AI chatbot today are designed
27:23
for single player. At Notion, we spent years building the
27:29
essential foundation for teamwork with features like real-time collaboration,
27:34
permissions, version control, and security guard rails.
27:39
Now we can extend this foundation to the agent and your notion workspace becomes
27:46
the place where humans and agents seamlessly collaborate.
27:54
Finally, a knowledge work agent needs to actually do real work for you. And this is the
28:02
most tricky part because when your tools are rigid and
28:09
your work context is fragmented, AI agent can work reliably.
28:14
They're just like people. No, they will struggle to piece everything together.
28:21
But notion has all the work context and notion has all the core building blocks
28:26
of knowledge work. So we sent notion AI to the boot camp
28:33
and taught her how to do real work using those building blocks and contact in notion.
28:40
With a lot of hard work and completely rebuilding Notion AI from the ground up
28:45
multiple times, we can say that it's finally ready.
28:56
And the result is anything you can do with notion, your
29:02
notion AI agent can do. is the state-of-the-art AI agent that can do
29:07
multi-step continuous autonomous work sometimes 20 minutes at a time.
29:14
This is a such a big step change in our company history and give us a brand new
29:19
foundation to build our future. So we want to finally turn a page.
29:28
Introducing Notion 3.0.
29:36
It's the It's the world's most advanced knowledge
29:42
work agent designed for teamwork. It's the first of its kind. We're really
29:48
proud of it and we can't wait for you to try it. Before Notion was your tool. With 3.0,
29:57
Notion is your AI teammate. Before notion was your notes app, your
30:03
knowledge base, your task tracker. With 3.0, notion can do work for you. It
30:11
can take your notes. It can give you the answer proactively. It can do your task
30:17
end to end. We think will really reduce your busy
30:22
work and give your time back for your life's work. Okay, let's see some demos.
30:32
And here's here to tell you more.
30:38
Wow, what a great intro.
30:43
Thank you, Ivan. Um, I'm Shear and I work on AI at Notion. I've been working
30:50
on AI for more than 10 years and it's never been more exciting than it is
30:55
today. But keeping up with AI comes with
31:01
this. I don't like it either. Uh so much of my
31:07
reserved focus time is spent on things like writing status updates, prepping
31:12
for meetings, updating docs. But recently, I've been able to reclaim
31:19
more of my focus time and actually focus because I've been delegating my busy
31:25
work to this guy. This is our brand new Notion agent. It's
31:33
being rebuilt from the ground up. It's new and improved, and it can tackle much more complicated tasks like this one you
31:40
see right here. Previously, Notion AI could only do
31:45
things one thing at a time. A single search, a single generation. You had to
31:51
give very precise instructions for it to do the right thing. Today, the agent is
31:56
able to take on a much more complicated goal or task just like the ones you see here.
32:04
We have taught this agent to do everything that you can do in notion and take advantage of the full set of
32:10
building blocks. It can generate toggles, bullets, checkboxes, and my
32:16
personal favorite, the table, columns. Anything you can do in notion, this
32:22
agent can do. We're also introducing instructions and
32:28
memory. This lets you control how your agent behaves.
32:34
It's also a place for you to store notes on your preferences and memories about
32:39
you. And since it's just a notion page, I can edit the instructions and memories
32:46
however I want and I connect it appention to the rest of the context in my in my notion workspace.
32:53
So instead of telling you, let's show you
33:00
In today's demo, I'm going to be a PM working on follow-ups from our offline
33:05
mode launch. First, I'm going to have my personal agent help with user feedback,
33:11
and then I'm going to have it write a product requirements doc or a PRD that I can share with the rest of the team.
33:17
Let's dive in. As I get situated in my day, I'm going to first start with a task of gathering
33:24
user feedback. I'm going to ask my agent to create a doc. Now, I know that there
33:30
is a bunch of information from Slack, from email, and from the web. So, I'm going to prompt the agent to look there.
33:36
I also want it to pull very specific user quotes from me. This is the kind of
33:42
thing that would have taken me hours to do all on my own. I would have had to comb through all of the user feedback,
33:49
create categories, copy and paste a bunch of quotes from everywhere. But this agent can do it in seconds. This is
33:57
our revamped Notion agent. It's got access to enterprise search, so we can
34:03
gather context from mail, from Slack, from Notion. With Notion's AI meeting
34:08
notes, I can pull quotes directly from transcripts with customer interviews. I can also search the web where probably
34:15
some of you have left us some feedback about offline mode. We hear you.
34:20
And because this agent can do anything I can do in notion, it can take all of that information and put it into a doc.
34:28
I just had one prompt, but this agent was able to take all of the different steps, string them together, and put
34:36
them into a research report. Let's check it out.
34:42
We can see here we've got quotes from everywhere and themes. I'm going to consider this done.
34:48
Next, we're going to write our PRD. So, again, I can ask the agent to get me
34:54
started. We're going to go back to the research report and use that as a jumping off point. From here, I can
35:00
prompt my agent again. I want it to use the feedback we just collected.
35:06
Again, this is something that manually would have taken me an hour or two reading through all of the user
35:12
feedback, extracting out the user problems, figuring out what's a feature request, what's a bug, organizing it
35:18
into a road map. But this agent here, it can do it in seconds.
35:26
Let's check it out. This PRD looks pretty good, but it's not exactly
35:33
in the style that I like to write and my team likes to use for PRD templates.
35:38
This is where instructions and memories come in. So, we're going to get it set up. To get it set up, we just click on
35:46
the agent and hit personalize. From here, I can add a new notion page.
35:52
And this is just a page. So it's going to work out of the box with all of your
35:57
existing setup. The flexibility that you have for organizing your thoughts in
36:03
notion, you can also use to organize your agents memory. We're going to add
36:08
our first memory here, which is that PRD template that I want to use. All I have
36:13
to do is add this here once. And now the agent can remember this for every future
36:18
interaction. We can put style guides in here. templates and really any preference I have. I'm also gonna ask it
36:26
to interact with me in a special way. I'm a little bit of a diva. So, I'm
36:31
gonna ask her to refer to me as a queen. And it's not personal without a
36:37
personality. So, let's give her some. All right, back to work. Let's go back
36:44
to our PRD and fix it up. I'm going to ask the agent to update this to the
36:50
right format. Now, I didn't add the PRD template here at all, but the agent was
36:57
able to pull the instructions, read the PRD template, look at it, review it, and
37:03
now it's uploading my updating my doc to the right format with the headers and the bullets that I need. Almost done,
37:10
but I really like to add a TLDDR section to the top of every single doc. So, I'm
37:16
going to ask the agent to do that. I'm also going to ask the agent to remember this for the future. Now, remember, my
37:23
instructions page is just a page. And since the agent can edit pages, the
37:29
agent can remember this all on its own. I never need to open the instructions.
37:35
Okay, I think we're done. Boom. Last step here, sharing with the team. I
37:41
think this PRD is ready for some feedback. So I'm going to share it with Yakob from our engineering team.
37:47
I think we can consider our busy work done.
37:56
So that is the brand new personalized
38:02
agent. It can reason in multiple steps. It's personalized to how I like to be
38:07
called and it's better each time I use it. For any PM researching docs, writing
38:14
docs, this can save you countless hours of busy work. And it's not just me.
38:20
We're also seeing fast growing companies like Forcel save 9 hours per person per
38:26
week. That's a full day back every week. We've been rolling this out slowly over
38:32
the last few weeks and this feels different. There's a lot of excitement.
38:38
And please keep the feedback coming. The good, the bad, the ugly. We want to hear
38:43
it. There's a ton to explore with your new agent. I only showed you a small
38:49
fraction of what the agent can do in my demo just now. It uses all of the latest
38:54
AI models. It's available on mobile and it's connected to all of your enterprise
39:00
connectors. But the best part,
39:05
everything I showed you, multi-step reasoning, enterprise search, personalization, all of it is available
39:11
today.
39:19
But I skipped over a pretty big Notion
39:24
building block. The real power of the agent comes where you combine it with databases. Here's Jacob to tell you
39:31
more.
39:38
Hi everyone, I'm Yakob from the engineering team and I'm excited to show you what's new with databases in Notion
39:45
3.0. In Shar's demo, we saw how the agent is able to search across many sources to
39:51
put together a high quality document. But what happens when you're working
39:57
with a huge amount of data? In a database, you might have hundreds or even thousands of pages.
40:04
For example, you might use the database to manage your product road map where each item needs regular updates.
40:12
Or for something like a task tracker. As you can see, there's a ton of detail that needs to be captured about every
40:17
task. The due date, the assigne, the priority, and then more details inside
40:23
of each page. You can use databases for so many things. I actually have a personal
40:28
database where I keep track of my favorite music albums and what I want to listen to next.
40:35
And I love that I can view my databases however makes sense for what I'm tracking.
40:40
Now, I love Notion databases, but even for a longtime Notion user like me,
40:46
building and maintaining large databases can be a lot of work. setting up the views, creating the properties,
40:53
populating the database with hundreds of rows. Sometimes this can feel slow or tedious. But luckily now, anything I can
41:01
do in Notion, my agent can do for me. And with databases, the agent can make me 100 times more efficient. But like
41:09
Ivan said earlier, the agent needs to be built for teamwork and it needs to respect your permissions.
41:16
And unfortunately, until recently, database permissions have always been all or nothing. Which means if I want to
41:24
share a database with someone like a client or an external contractor, I have to give them and their agent access to
41:30
the entire database or they get no access at all. This has been a major pain point for
41:37
years. But today, I'm excited to announce that we're launching granular database permissions.
41:51
It's our number two feature request of all time. Second only to offline mode.
41:58
This feature gives you more control over how you work with your teammates and their AI agents. You can now set up a
42:05
database so that a page is automatically shared with whoever it's assigned to or
42:10
whoever is mentioned in a particular property. I have a lot to show you. So, I'm going to show you in a demo. In the
42:17
demo, we're going to do a few things. I'm going to first have my agent turn shears PRD into specific tasks.
42:25
And then I'll have the agent look through Slack to find bug reports.
42:31
And finally, my agent will help me with adding views and formulas. So, let's
42:36
jump in. If you're an engineering lead like me,
42:42
you spend your time focused on two things. Building new features or fixing and maintaining what already exists. And
42:49
when you're building new features, one of the more timeconuming parts is breaking down a product spec into
42:55
smaller tasks. Normally, you'd have to manually move a lot of information from one place to another. But now, my agent
43:03
can do that for me. So I just asked for it to break down this PRD
43:08
and I already have this database set up with where our team is tracking our work. So the agent is simply adding the
43:15
tasks here with all of the necessary properties populated. Without AI, I'd need to read the PRD,
43:23
add names to the tasks, assign owners and priority levels. But with AI, that
43:29
just happens in one pass. And I love how much time this saves me so that I can focus on doing the things
43:35
that I really enjoy like writing code and working with my team to solve hard problems. So as you can see the database
43:42
is now fully populated. All of the properties, the assenee, the task type, priority. All it took was one prompt and
43:50
now we're ready to move on to the next thing.
43:58
Now, let's talk about how my agent can help me with the second part of my job,
44:04
fixing bugs and maintaining existing features. As Sher mentioned earlier, I can
44:10
personalize my agent with instructions that make it effective at performing my tasks. So, I taught my agent to mount
44:18
route tasks to the right owners because each task on my team falls into one of
44:23
three categories. front-end tasks, back-end tasks, or mobile tasks. So, as
44:29
you can see here, I have this mapping in my agents instructions. So, it knows who
44:34
should be assigned to what. And then, importantly, when I chat with the agent,
44:40
it's able to use that mapping automatically. The flexibility of the agent's memory
44:46
gives it so much power. So, now we'll see it in action.
44:52
Now, we're going to ask our agent to consolidate feedback and bug reports into notion. I also want to make sure it
44:58
assigns owners and due dates to make sure that we stay on track with our work and our timelines.
45:04
Without agent, I'd have to copy each of what each of these one by one and it will probably take about an hour or two
45:10
because there are more than 30 reports in our Slack channel. But like I said before, in databases, my
45:19
agent can make me so much more efficient. And I love that the agent is
45:24
able to read data from all the places where teams do work. It can search your
45:29
email in notion mail or Gmail. It can also meet you where your teams chat like
45:34
Slack or Microsoft Teams. It's also able to research on the web.
45:41
As you can see, it's creating tasks, dduping them. You can see it pulled items from Slack and it shows each step
45:47
so I can trust the results. Finally, if you look at this sliced view by assigne, all of the tasks were routed to the
45:54
right people and the mobile tasks, the front-end tasks, and the backend tasks
45:59
were all mapped to the people who I suggested and it was just like I specified in my agents instructions.
46:10
Thank you. Let's now look at how we can use Notion
46:16
Agent to help me track the balance of work across my team.
46:23
I need to find a way to visualize how many tasks each person has assigned to them. So, I'm just going to ask it to
46:30
add a bar chart and I want it to be sliced by assigne. If you're a Notion power user like I am,
46:37
creating views is something you're already comfortable with, but I personally prefer to just describe what
46:43
I want and let the agent take over instead of manually needing to create this from scratch.
46:50
There it tells you it's done. Now we're going to navigate and look at the chart. And now we're able to see an easy slice
46:56
of tasks and the distribution of work at a glance. So then I can use this to load
47:02
balance work across my team. Okay, before we move on, I want to know, do we have any formula fans in the
47:08
audience? Amazing. Um, formulas are really powerful, but they can also be
47:15
intimidating. So, I need to find a way to see how much
47:20
time we have before each task's due date. So, we'll add a formula for that.
47:26
We'll name it days until the due date. We'll pick the property type here. And
47:31
then importantly, instead of writing the formula from scratch, we're just going
47:36
to have AI generate it for us. We're going to say number of days until the due date. And I wanted to add some color
47:43
so that we can easily tell which tasks are coming up soon and which ones are overdue that we need to get to. If you
47:49
haven't used formulas before, or even if you have, using AI can really help you build anything you want. So, after some
47:57
thinking, there it is.
48:05
The formula looks correct and we're able to use the colors to be able to tell which ones are coming up soon. We have
48:11
the ones in red are overdue, the yellow ones coming up soon, and the green ones we have more time. Finally, there's one
48:17
more thing that I want to get right. I want to make some changes to the permissions of this database.
48:24
So, I'm working with some contractors on a few of our tasks, and I only want them to see the tasks that they're assigned
48:30
to. So, I'll set a permission rule so that anyone in the assigne property can
48:36
get canedit permissions to their tasks and nothing else. So, now our contractors can see their work, but they
48:43
can't see anyone else's.
48:51
Okay, that's plenty for today. Now, that's just a small taste of what you
48:57
can do when you combine your personal agent with Notion databases. It's incredible to see the perfect database
49:04
created and modified for you like it has for the browser company CEO Josh and so
49:09
many customers already. And we'll keep making databases better with new features like conditional coloring,
49:16
which lets you highlight each page based on a particular property. like if you want your high priority tasks to be
49:23
marked in red like they are here. This is available now so you should give it a try if you haven't already.
49:30
And we're also adding a new type of database view map view.
49:37
So you can use this to easily easily visualize your data plotted on a map.
49:43
Stay tuned for this coming very soon.
49:55
Databases offer an incredible toolbox for your personal agent to work with, and I think they truly help your
50:02
personal agent build beautiful tools for you. Our new database permissions and
50:08
everything I showed you with agents and databases is available starting today.
50:18
They're included in our business and our enterprise plans and they're free to try for anyone. So, we've seen how Notion AI
50:26
can complete work for you in Notion Docs and Notion databases, but we know Notion isn't the only tool you use, which is
50:33
why I'm excited to invite MJ to the stage to show you how the agent works with your integrations.
50:43
Great. Thank you. Hi everyone, I'm MJ and I'm from the product team. So far, we showed you that
50:51
agents work super well with pages and databases. And now I'll show you two new
50:56
ways to connect notion with your other tools. Even though plenty of work happens in
51:03
notion, a few things might need to live elsewhere. For example,
51:08
if I'm vi coding or prototyping, I need to use a separate tool like cursor, but
51:15
all of my context is in notion. And it's so much busy work to keep switching in
51:21
between tools. Back and forth, copy, paste, command, tap, repeat.
51:29
But that's where the notion MCP integration comes in. It's a secure way
51:35
to read and write to and from notion with other tools. So all of your context
51:42
from notion so you can use all of your context from notion without being in notion.
51:51
And today I'm excited to announce that we're making the MCP even better by
51:57
adding support to all of your tools from enterprise search. That means while
52:04
you're vip coding in tools like cursor, you'll be able to reference things like Slack conversations, GitHub pull
52:10
requests, and much more.
52:18
And we're also adding new integrations for your personal agents to connect to Notion Calendar and Notion Mail. This
52:27
will make it super easy to schedule meetings and send emails, way faster than human speed.
52:35
So, let's see it in action. First, I'm going to use the notion MCP to build a
52:40
prototype in cursor by pulling context directly from a notion database.
52:46
And once that's done, I'll use the MCP again to update that same task from cursor.
52:53
And finally, with my notion calendar and notion mail integrations, I'll use my
52:58
personal agent to set up a few customer meetings to get some feedback on our prototype.
53:05
All right, so let's start with this task to add more information to the offline settings table. This will help users
53:12
understand that they've downloaded the right pages before going offline. All I need to do is ask Cursor to
53:18
implement this task. And within a few seconds, cursor is going to realize that
53:24
the the details it needs lives in notion. And it uses our MCP to bring in
53:30
all of that context seamlessly. From our task, it learns that what it
53:36
needs to build and starts building our prototype. Normally, I'd write some tests, but in
53:43
this case, let's watch the code changes come in live in our prototyping environment and confirm that everything
53:49
looks good. So, we're looking for the last viewed, last synced, and modified by columns.
53:58
This is looking pretty good. All right. So, next, we'll use the MCP again to
54:04
update the task from cursor to reflect all of our hard work. And just for the
54:09
demo, let's jump back to notion so that we can watch the task move from not started to done without me touching
54:15
anything in notion. Now imagine how useful this is for
54:21
engineers as they're hacking away in their favorite IDE. They now can update notion without switching between tools.
54:28
Now that we have a prototype, I want to chat with some customers to get some feedback. I have a list of all of my customers and
54:36
their email addresses, but it can be really tedious to follow up with them and schedule those sessions. So, let's
54:43
have my personal agent handle that for me. My agent finds and uses helpful
54:51
information in Notion to learn more about each customer and their needs.
54:58
Then it pulls in contacts from notion calendar and picks some meeting options based on my working hours, existing
55:06
meetings, and buffer time preferences. And if you're a PM like me, you probably
55:12
can relate to the feeling of having negative time between your meetings and feeling like you're just perpetually
55:18
late. Luckily, my personal agent handled all
55:23
of those headaches for me with these amazing recommendations. So, in a few
55:28
sec seconds, the agent is going to create the invites, then send them to me and each customer.
55:38
So, let's quickly hop into my notion calendar to watch the invites appear.
55:45
And boom, we have the events scheduled. and I've been saved from so much busy
55:52
work. Next, I'll ask my personal agent to send a couple of follow-up emails to our
55:58
customers. Agent remembers all of the emails, so I don't have to copy and paste or even
56:06
write the emails myself. I love that personal agents take the
56:11
guesswork out for me. I don't need to worry about crafting the right customerf facing tone or accidentally sending out
56:18
a typo. It tailors each email to each customer, including their name, the
56:24
respective invite details, and a clear agenda based on each customer's feedback.
56:30
So, the drafts are done, the emails are being sent, and the magical part is that
56:36
with the same amount of effort, I can send out countless more emails to additional customers.
56:42
Calling this just a timesaver is an absolute understatement.
56:56
So we just used the notion MCP to build a prototype and follow up with
57:02
customers. And what's extra exciting is that you don't need to be an engineer or a developer to use the MCP. There are no
57:10
API keys and it's super easy to connect and the MCP is so powerful that even the
57:17
cursor team themselves are loving it. I'm excited to share that we have
57:23
expanded our list of firstparty MCP integrations with lots of new partners
57:29
including lovable HubSpot Windsurf and many more.
57:34
The not the notion ecosystem is even more open than ever before.
57:40
You're free to combine the flexibility of Notion's building blocks with whatever tools you need to do your best
57:46
work. The MCP is also available today and free
57:51
for everyone.
58:00
And if you're on the business or enterprise plan, you can use the MCP with advanced capabilities like
58:06
enterprise search with AI connectors. We have a lot more planned here, so please stay tuned for more improvements.
58:14
And with that, I'd like to invite Erica to the stage to tell us some exciting updates for the Notion community.
58:22
All right. Thank you, MJ. And a big thank you to
58:28
Yakob and Shear. Let's give it up for those incredible demos one more time.
58:36
Hello everyone. I'm Erica Anderson and I support our global go-to market teams here at Notion. As Ivan mentioned
58:44
earlier, Notion has always been built for teamwork. It's the place where your docs, your databases, and your projects
58:51
come together so your teams can get things done all in one workspace. And
58:58
today we saw powerful demos showing a new way of working. One where you can
59:04
assign your task to your agents, free up your time, and focus on the work that
59:09
matters the most. And if you're thinking, "This looks really amazing, but where do we start?" You do not have
59:16
to figure it out alone. That's where our incredible community comes in. These are
59:22
folks all around the globe who specialize in notion experts in all the
59:27
latest AI capabilities and passionate about helping teams create better systems for their day-to-day work. From
59:36
ambassadors to developers, template creators to consultants, our community
59:41
has the tools, the teams, the resources you need to get the very most out of
59:46
your new agents in Notion. which is why we're very excited to share
59:52
a major update. We have completely re-imagined our consulting partner
59:58
program. And our goal is very simple. It is to make it easier than ever for you
1:00:03
to find the perfect partner to help you implement notion 3.0.
1:00:09
So take a firm here. With over 2,000 employees, they needed a single
1:00:14
operating system to bring everyone together. So they worked with notion state, one of our partners to build a
1:00:20
notion workspace for all their company knowledge, all of their workflows with AI at its core. Now building the system
1:00:29
is the very first step. Our consulting partners go beyond building. They help
1:00:34
teams adopt new ways of working and make it stick. And for a firm, 95% of
1:00:41
employees adopted notion. And they even cancelled some SAS subscriptions along the way,
1:00:48
consolidating into one AI workspace, saving money, and helping their teams
1:00:53
move faster. To make it simple to connect with partners like Notion State, we're
1:00:59
updating our marketplace today. Now, in just a few clicks, you can browse certified consultants, see their
1:01:06
specialties, and find the partner who's right for you. And if you're sitting here wondering, could I do this, too?
1:01:14
The answer is absolutely yes. If you're passionate about Notion, you can launch
1:01:20
your own consulting business. Head over to Notion Academy where you'll find new coursework, certifications, and badges
1:01:27
that can show your expertise. And as a certified Notion consultant,
1:01:32
you'll gain exclusive benefits like new leads, priority support, and more
1:01:38
co-branding. And unlike other platforms that take a cut, with Notion consulting,
1:01:44
you keep 100% of your revenue.
1:01:52
And consulting is just one of the many ways to build a business with Notion. In
1:01:58
the Notion marketplace, you'll see the community offering templates for every imaginable use case. Today, there are
1:02:05
over 50,000 templates in Marketplace. And now we're adding something brand
1:02:11
new. Yes, agent personalization templates. As
1:02:17
you saw in the demos, agents become even more powerful when you give them instructions about how you work. And
1:02:24
naturally, our community is already running wild with this idea. Like this
1:02:29
one from Lou. These agent instructions make sure that Notion AI always writes
1:02:34
in your brand's tone and voice. They even include a glossery so it always gets the terminology right. Your team
1:02:41
gets onbrand writing from that very first draft. And that's just one way to
1:02:47
personalize your agent. Take this template from Dagmar. It turns your
1:02:52
personal agent into a friendly advisor that helps you focus on your goals. It's
1:02:57
grounded in productivity frameworks like getting things done and atomic habits
1:03:02
and teaches your agent exactly how you prefer to organize your notes and your tasks.
1:03:08
Think of Marketplace as the app store for your agents personality. Whether you
1:03:14
want it optimized for productivity, built to reflect your own unique style, or designed to bring a little fun like
1:03:21
calling you queen, you'll find it in the marketplace. It's a whole new way to
1:03:27
make AI feel like it truly works for you. There's so much creativity in what
1:03:33
people are building already, and it is just the beginning. With that, let's
1:03:38
take a quick look at what the community and customers have been reacting to with
1:03:43
all of these new announcements. In Notion, it's easy to mimic the
1:03:49
structure you have in your head in the computer. It just feels like a natural extension of the way that you might
1:03:55
think. With Notion, every individual person at RAMP has an AI assistant that helps them do their job.
1:04:00
It's a co-orker that's been around that really knows and has genuine context for why we've made certain decisions.
1:04:06
6 minutes later, I'm looking at a really informed graph, chart, however I want to visualize that information. Before
1:04:12
Notion, that would have taken me 6 8 hours. We felt the benefits of AI in
1:04:17
notion as we've gone from a few people to over 100 people very quickly.
1:04:27
This is quite transformative and it's not the half of what it can do. Notion AI agents will just further
1:04:32
evolve notion to be the onestop tool. This changes everything.
1:04:46
Thank you, Erica. So, today we ship a lot of things for
1:04:52
you, including the state-of-the-art knowledge work agents with multi-step
1:04:57
actions and memories. We launched our top request features,
1:05:03
database role permissions, and a brand new group of MCP and AI
1:05:09
connectors. So now notion 3.0 really feels like
1:05:14
something new. But before I let you go,
1:05:21
we have one more thing for you today.
1:05:27
You know in our demos you saw how you can start delegate your busy work to
1:05:34
your personal agent. But knowledge work is a team sport.
1:05:41
And a lot of this repetitive busy work happens on a team level.
1:05:48
So we're thinking how can we create a new type of AI agent that automate
1:05:54
repetitive task for your team and allow you to create as many such agents as
1:06:00
your company needs. Agents that can work while you're
1:06:06
sleeping. So, here's Ryan to tell you more.
1:06:17
Thank you, Ivan. Hey everyone, I'm Ryan. I'm on the AI team here at Notion, and I
1:06:24
cannot wait to show you what we've been building. We saw many examples today of what your
1:06:31
personal agent can do. And it can do a lot. It's like a Swiss Army knife for
1:06:38
your productivity. But sometimes you need a specialized tool, something that's really good at a
1:06:46
specific job. And sometimes there's just work that you want to put on autopilot.
1:06:53
And so with Notion 3.0, that is now possible with custom agents.
1:07:02
Imagine a whole team of agents, each with their
1:07:08
own specialty, running in the background. So, you might be asking, "What are
1:07:14
custom agents?" Well, while your personal agent knows
1:07:20
everything that you know, with custom agents, you control the information that
1:07:26
they have access to. For example, you can give your custom agent access to a
1:07:32
specific database. That way, they limit their search results and keep everything focused and
1:07:39
trustworthy. Next, your personal agent only
1:07:45
collaborates with you, but with custom agents.
1:07:50
You can share them across your entire team, supercharging what everyone can do
1:07:56
in a day. And finally, your personal agent only knows to work
1:08:04
when you use it. But with custom agents, they work autonomously and in the
1:08:11
background. They can take action based on triggers or even schedules. Just
1:08:17
imagine waking up each day with your new daily briefing delivered by your custom
1:08:22
agent. You can create as many custom agents as
1:08:28
you need. They can specialize them to any role or team from employee
1:08:34
onboarding to inventory management and everything in between.
1:08:41
Now that is enough talking. How about I actually show you how they work?
1:08:48
Here is an example Slack feedback channel the offline team uses to collect feedback.
1:08:56
They're spending a ton of time answering the same question over and over and
1:09:01
over. Now imagine I run it for our company and I've been tinkering on a
1:09:07
custom agent to help them out. Let's see it in action. I'm going to go ahead and
1:09:13
ask when did offline mode ship? If I wait a second, I've got a new
1:09:19
reply. Meet Offline Oliver, my team's custom
1:09:24
agent that can answer questions about offline mode. And yes, it did ship on the August 12th and we announced it on
1:09:31
the 19th. But you don't just have to use Slack to work with custom agents. Each custom
1:09:38
agent comes with its very own chat where you can ask it questions like, "Do
1:09:43
databases work with offline mode?" And the answer of course is yes.
1:09:50
Now let's see how custom agents work under the hood. So I'll go into offline
1:09:57
Oliver settings. And up at the top we have our triggers.
1:10:03
These are what wake up your agent and get it to work. You can set triggers like a custom schedule. It can respond
1:10:10
to events in notion or even external events in Slack with many more apps and
1:10:16
integrations coming soon. Below our triggers, we have our instructions. This is where I can tell
1:10:24
our agent exactly how it should work. Like in this example, reading from a knowledge base.
1:10:31
And last, we have our tools and access. This is where I can decide exactly what
1:10:37
my agent can see and do within my workspace.
1:10:42
So, offline Oliver is already pretty useful, but it's just answering questions.
1:10:49
What if our agent could file tasks automatically? Well, our team uses this people database
1:10:56
to keep track of who works on what? We can give offline Oliver access to this
1:11:02
database to automatically assign new tasks.
1:11:07
Now, I could go into our custom agent and type some new instructions, click
1:11:12
some buttons, or I could just ask Notion AI to do it
1:11:17
for me. So here I'm going to tell notion AI to use the task database, the people
1:11:24
database, how to assign people, and even how to handle replies in Slack. And
1:11:30
Notion AI will figure out exactly what updates to make to our agent.
1:11:37
And just like that, our agent's updated with new
1:11:42
instructions and configuration ready to use. All right, let's see it in action.
1:11:50
I'm going to pop back over to Slack and ask a question about some weird Android
1:11:55
bug that I've been seeing. I wait a minute and now we have another reply.
1:12:02
And who would have guessed? It's offline Oliver again. This time with a new task
1:12:07
automatically assigned to Yakob. But Yakob or MJ is the one that typically
1:12:14
helps us with Android. So without leaving Slack, I'll just tell offline
1:12:19
Oliver to make some updates to the task and most importantly the people database.
1:12:26
Now if I go over to our people database, I see Android listed under MJ's profile.
1:12:31
So future tasks will go straight to her. What we've just built is a selfupdating
1:12:38
agent that can learn on the fly and update its own memory. And this is just
1:12:43
a glimpse of the power of combining custom agents and notion blocks. Let's meet another agent, Grug. Grug's
1:12:52
an honorary member of our design team. At this point, we give Grugs and Grug
1:12:58
spits back critiques. What about uh an agent for personal meal
1:13:04
tracking? Well, I built that, too. I send it pictures of what I ate. It can
1:13:10
analyze our images and automatically log everything into my personal meal tracker.
1:13:17
Now, the last one, we spend a ton of time collecting feedback from folks like
1:13:23
you. So, we build an agent to help us do the busy work of collecting the feedback. This agent goes to the
1:13:30
internet every morning and turns what people are saying about offline mode into new reports with feature requests,
1:13:37
sentiment analysis, even filing bugs ready for us to work on.
1:13:44
This one agent can save our researchers and product teams hours every single
1:13:51
week. We gave early access to some of our
1:13:56
friends at RAMP and they're telling us how much it's already helping them get work done. And you're going to see some
1:14:03
more exciting examples from Ben in just a few minutes. Custom agents are already in early
1:14:10
access and will be shipping more broadly soon. We cannot wait to share these with
1:14:15
you and see what you build. And if you want to see a deep dive into how we use custom agents at Notion, please join us
1:14:22
later today on stage two. And with that, let's see a quick video of everything
1:14:28
new in Notion 3.0.
1:15:34
online. Thank you so much.
1:16:13
I hope I hope you like that video as much as we did. Our team really had fun
1:16:21
with that.
1:16:27
So today we show you that you can use notion agent to delegate your busy work
1:16:34
and get time back for your life's work. We also show that teams can augment
1:16:40
their capability with custom agents. So a team of three might soon have the
1:16:46
output of 300. All these bring us back to our mission
1:16:52
to help you build beautiful tools for your life's work. For the rest of today,
1:16:59
please don't be shy to share your feedback with us. You know, we're easy to find because we're going to wear the
1:17:04
Notion jackets. Thank you for joining us today and have a great make with Notion.
1:24:38
Please welcome Welcome Ben Levik.
1:24:45
Hello. Hello. Hello. Make with Notion. Uh, anybody else excited about Notion 3.0? Yeah,
1:24:52
it's pretty good. Pretty good. Well, my name is uh my name is Ben Levik. I work at a company called RAMP and I lead
1:24:58
operations and some of the AI product teams there. And I'm really excited to uh follow up that amazing keynote with
1:25:06
some of the lessons we've been learning at RAMP about how to bring AI fully into your team. I think uh notion 3.0 is
1:25:13
something we've already been spending a lot of time exploring with our team and I hope to bring a mix of kind of the
1:25:18
practical and the aspirational of what we're doing. So to really understand kind of why I'm here, uh you know, RAMP
1:25:25
has been interested in AI since it was founded, but we started to notice about a year ago that there was an
1:25:31
acceleration in what AI could do for business. I imagine many of you have noticed the same thing. And so we
1:25:36
started to get really serious when the executive team got together at the beginning of the year. We said, we probably need to really get focused on
1:25:43
AI for our team. And so we set what I think was like a humble achievable mission, which was to become the most
1:25:49
productive company in the world. And uh yeah, you might say RAMP likes to bite off more than it can chew. We do
1:25:55
that sometimes. But to understand why I actually think this is the right mission for our team, it's probably worth
1:26:01
spending a minute on what RAMP does and why we exist. Ivan mentioned this a little bit in his keynote, but RAMP's
1:26:07
external mission is to save our company, our customers their most valuable resources, their time and their money.
1:26:14
Time is money. Uh, and so we're really focused on bringing a set of finance tools to customers that are going to
1:26:21
really unlock their most precious resources, keep them focused on their business.
1:26:26
On the outside, that takes the form of a suite of powerful financial tools, everything from corporate cards to
1:26:32
expense management to bill pay. And they're all for both finance teams, but also anybody who spends money in a
1:26:38
company. And so if any of you are dreading going home after make with notion and filing an expense report for
1:26:44
an hour and a half and digging through your bag for receipts, RAMP is there for you. If any of you have ever gotten into
1:26:50
an infinite loop with your finance team and boss about whether you can get a new vendor approved, RAMP is there for you.
1:26:56
And if any of you run your own business and spend way more time than you should moving money between different bank accounts just so that it's where it
1:27:02
needs to be, RAMP is there for you, too. We really are focused on putting money in the right places and saving you time
1:27:08
doing it. And so AI agents unsurprisingly has been a big focus of ramp as well. Under the hood, we really
1:27:16
believe that AI agents can drive powerful new behaviors within finance.
1:27:21
And like the notion team mentioned, uh there are AI agentic behaviors that we
1:27:26
can put everywhere. We've launched a set of agents already uh this year. We're actually going to launch more next week at our conference. So Jinx to notion
1:27:33
agents is in the air and we believe that they should be there to help humans focus their time on making fewer more
1:27:39
critical judgment calls. Let the robots do the busy work. Uh agents can do the uh and humans can do the rest. And when
1:27:46
you have this productivity mission, every day that goes by culturally as a company feels like a burning need to do
1:27:52
more because you see all of your customers, you see how they are wasting their time and their money and you want to go as fast as possible. So velocity,
1:28:00
customer obsession, these are the values that power ramp. And it's given us the ability to serve 45,000 customers in
1:28:06
counting for which we've saved almost 30 million hours of wasted time, $10 billion of wasted money. And it's given
1:28:12
us the opportunity to partner with like-minded companies like Notion. So we rolled out Notion at the beginning of this year. Notion also runs their
1:28:19
finance on ramp. So it's two productivity companies productivitying each other's productivity. It's a great
1:28:24
productivity dream. So if we get back to the internal mission, uh to become the most
1:28:29
productive company in the world. Okay, uh that's the why, but what is the how?
1:28:35
Obviously, AI agents are going to play into this. Uh but it's a pretty simple question to answer how AI agents are
1:28:40
going to make us productive. It's by making each team member the most productive possible version of their work.
1:28:47
In some ways, we do that by making teams faster. Uh AI can automate a lot of manual processes. You just saw 17 demos
1:28:54
of how that can be done. It can instantly deliver information where it needs to be. Importantly, it can also
1:28:59
make you better at your job. Uh there's also a lot of ways I've seen AI make people worse at their job. So, be careful here. But to become better at
1:29:06
your job, it's really using the deep research capabilities, the ability of AI to go further than you can into
1:29:12
understanding things and bring you back what's important so you can make the best possible decisions.
1:29:17
But the real magic, the rubber hits the road is that AI is faster and better at
1:29:23
the same time. When you can do all of that deeper research instantly, then you can afford the opportunity to make
1:29:30
bigger decisions with more information. When you make bigger decisions with better information, you can afford to go
1:29:35
faster as a team. And so we've really tried to find is those opportunities to get our teams faster and better
1:29:40
simultaneously. Where do we see that happen? Uh notion mentioned this too. Coding is an obvious one. We have seen
1:29:48
tools like cursor uh cloud codeex really power up our engineering team. It's
1:29:53
gotten us to ship 270 features just in the first half of this year, which is more than one feature a day. It's more
1:30:00
than the entire year in 2024. The engineering team is running really fast. They are uh deploying a DevX team to
1:30:07
just teach them new tools. And so the impact's really clear there. What I think is a question is what about
1:30:14
everything else? We just saw a lot of amazing tools from Notion, but how do you actually put them into practice? And
1:30:20
I think here I talk to a lot of people on my team and a lot of colleagues in other companies and what I often see is
1:30:26
that there's a few limiting beliefs that are often slowing teams down in their ability to really get under the hood of
1:30:31
how to use AI within a team. I'll go through three of them. The first one I'm going to call the doomers. This is the
1:30:37
folks who basically say, "Well, AI is going to take all of our jobs. We should probably just roll over and wait for it
1:30:43
to happen." Okay. Uh it's an interesting take. It's one that if it comes true will be
1:30:49
unfortunate, but I think it's a it's a reason not to take action when action is actually the most important thing to do.
1:30:55
The second one I'll call the zoomers. They're kind of on the other side of the spectrum, which is agents are becoming
1:31:01
or AI tools are becoming more powerful every month. LLMs are going wild. In a few months, we can just buy a tool, hit
1:31:08
a button, and it'll do our jobs for us perfectly. Okay, I really hope that happens. That
1:31:13
would be great for my job. But I do think that it's going to be a little harder than this because that ignores understanding all the work that's
1:31:19
actually happening that needs to be replaced. And then the third one, I really just had to come up with another rhyming word, so I went with boomers. Uh
1:31:26
this is the one where it's people are saying, "Dang, AI is moving really fast. Every morning the news is different. I'm
1:31:32
going to wait a year or two because by then they'll have figured out the right way to do it and then we buy AI. Then we
1:31:38
get going." And I think this is probably the most dangerous one because it puts you in this passive waiting mode and I
1:31:43
guarantee you're going to be waiting forever where what you need to be doing is learning, practicing, getting the hang of how to do AI in this era. So
1:31:51
what I want to propose and what we've been doing at ramp this year is a fourth mode and I think it's going to get good reception and an audience like make with
1:31:58
notion and it's the idea of builders. It's the idea that work is changing
1:32:03
fundamentally. the way that you do your job today or that you did it a year ago should not be the job that you're doing
1:32:09
in a year or two. But if you're curious and if you're driven, you can become a builder that's going to usher in the new
1:32:15
version of your job. I think the nice thing about builders is it puts you back in the driver's seat in the AI era.
1:32:22
And in some sense, it also brings us back to something more primal. I've shamelessly stolen this image from the first Make with Notion conference
1:32:28
because I love it so much. But we all started as builders. We all got into Legos and blocks and puzzles. And then
1:32:35
somewhere along the way, knowledge work became something other than building. It became you stuck in a set of tools that
1:32:41
you have to use exactly some way doing the same work day after day. And I think what AI can unlock for every person in a
1:32:48
company, not just the engineers, not just the coders, is a sense that you are now back in the driver's seat to build
1:32:53
the type of work you want to do. This is obviously something that we get a lot of help from notion with. And the way we've
1:32:59
gone about delivering it is in three stages that I'll title as kind of the chapters of prompt, knowledge, and
1:33:05
workflow. What do I mean by those three words? The prompt is getting your team or
1:33:10
yourself to be able to define precisely what you want AI to do. The knowledge is
1:33:16
to understand what that AI then needs to know, what data, what information in order to do the thing you want. And if
1:33:23
you've defined that well, it's probably producing something of value. in the workflow is then how do I put it where
1:33:28
and when I want it to be and if these three things feel familiar from the notion keynote and as I go through each
1:33:35
of them that's no accident uh one of the reasons that we're happy notion customers is that notion has been a key
1:33:40
part of each of those journeys for us so let's jump in let's talk a little bit first about what I mean by prompt
1:33:47
this is not the hundth LinkedIn uh advertised prompt engineering workshop
1:33:53
where I tell you that if you tell AI it's a genius it'll give you better results. Put that aside. When I mean
1:33:58
prompt, I mean getting your team or yourself in the driver's seat doing prompting with AI. What we did at RAMP
1:34:05
at the beginning of this year when we started this become the most productive company in the world journey is we gave
1:34:11
everybody ubiquitous access to a lot of AI tools. We more or less have a policy where if you request an AI tool that
1:34:18
involves you writing prompts for AI, we will say yes. But we specifically preload every laptop at RAMP with three
1:34:25
tools. ChatJBT, Notion, and Notion AI and Perplexity. And the reason that we
1:34:30
love those three tools is they force the person using them to write actual prompts that will have AI go and do
1:34:36
something. You don't get to just hit a button and then under the hood somebody else has decided what AI does. It means
1:34:42
you yourself have to be the one to define what happens. And we've had pretty good uh adoption on that. We have
1:34:49
about 1,200 people at RAMP. We have about 90% monthly active usage of notion and notion AI. About a 100% weekly
1:34:56
active usage of Chad GBT across the team. That's tens of thousands hundreds of thousands of individual practice
1:35:03
sessions every month that our team is getting. And I think that's really crucial. What I love most is this nine
1:35:08
stat out of the top 10 power users of chat GBT at the company. They represent nine different departments across the
1:35:15
team. So this is not just engineers. It's not just product people. It's operators. It's sellers. It's marketers,
1:35:21
it's uh it's everyone. And so getting ubiquitous adoption means we're getting those practice reps. And we've had to do
1:35:27
work beyond just giving you a license. It's not as simple as just here's chatbt, go have fun and do AI. We do a
1:35:33
ton of enablement. Uh every group that starts at RAMP gets kind of a onboarding boot camp for how we use these tools. We
1:35:40
update that every two weeks because it's always changing. We try to get you brought in at the cutting edge of how we
1:35:46
think you should be doing it. You separately go every quarter or so around every department in the company, hear
1:35:51
them out on the biggest problems they're having, and then try to find ways to get them started using AI to solve one or
1:35:57
two of those problems. Build the muscle. And finally, we have a little Slack channel. Ramp uses AI. We launched it
1:36:03
without doing a ton of marketing. Nobody's like forced into this channel. We're up to about 600 people, half the
1:36:09
company in the channel now. and it gets daily active usage just trading tips saying here I'm stuck uh and kind of
1:36:15
making the kind of wheels turn across AI and prompting in the company I'm on part
1:36:21
of this channel and and people separately often DM me on Slack and say Ben I just tried to use AI for something
1:36:26
and it was terrible I don't think AI works and what I ask them is what did you prompt it to do and too often what I
1:36:34
end up seeing is well I prompted it what are the benefits of notion for ramp And then I got a result and read this.
1:36:42
It's total it's not very good. It's very vapid and it doesn't say anything specific. It's a super vague response.
1:36:48
Why doesn't AI work? And you know, it's not bad AI, it's bad usage of AI. And so
1:36:55
it's a great lesson to be able to teach people. We have this epidemic of vagueness, I think, in how we do work
1:37:01
now. And while I think Google search is one of the greatest products ever made, I do lay some of the blame for it at
1:37:07
Google search's feet because if you're sitting in your cubicle and you're typing three and a half paragraph long
1:37:12
Google searches, your co-workers are going to peer over and wonder what's wrong with you. We all learned that the pattern for getting stuff out of the
1:37:18
internet is type three words, scroll around, click some buttons, eventually you'll find what you want. And the irony
1:37:25
is that AI demands exactly the opposite of us. AI is fil so focused entirely on
1:37:31
precision. It's a word I use all the time at RAMP now because if you can provide deep precise contextrich
1:37:37
information to AI it will usually deliver for you. I think it's easy to say so go be
1:37:44
precise you'll be fine. Uh but it's a lot harder to actually do it. I get writer block all the time when I stare
1:37:49
down this the scary chat GBT new prompt. I don't know exactly what I want. And this is where the biggest unlock I'll
1:37:56
share today really comes in. And that is to use AI to do better AI. You are not
1:38:01
alone in the journey anymore. You don't have to wait until your boss has a moment to help you. You have this all
1:38:06
powerful tool next to you that can help you out of the vagueness trap and into precision. And I call this the prompt
1:38:12
feedback loop, which is you start with your vague prompt. That's okay. Put it in. But rather than just hit enter and
1:38:18
get that terrible output, ask AI to ask you smart questions that will give it more information to work for you and
1:38:25
it'll give you a couple usually pretty good questions and now your job is to write a lot of stuff back to it. Uh,
1:38:30
this is literally AI extracting precision from you so that it can do better work. And you don't actually have
1:38:36
to write the follow-up prompt yourself. You just take all the stuff you splat it out and ask AI to summarize it into a
1:38:41
new and better prompt. And you can go around this loop as many times as you want. And when you feel like you're ready, you hit enter. You see if you got
1:38:47
the magic that you wanted. And if you don't, then you just go around the loop. Again, this is a pattern I am using now
1:38:53
every day and it is getting me much better results. I did two rounds of it with the, you know, how is notion good
1:38:59
for ramp. Ended up with a six-page summary. It cites a million quotes. It has a bunch of sources. This is the kind
1:39:05
of thing I send to my boss and get promoted. Like it didn't take a whole lot longer, but it just forced me to be precise.
1:39:12
There's two other pro tips I'll share on precision. Uh, one they they talked about it. AI meeting notes is amazing.
1:39:18
Getting transcripts of every conversation you have is amazing. If you're not doing it already, find a tool
1:39:24
and do it. I think that notion calendar along with notion I meeting notes is great. There are others, but one way or another, insert all of that context into
1:39:32
the AI world that you live in and use it to build precision. Uh what you can also
1:39:37
do now that notion agent is live is you can take say your weekly meeting you can
1:39:42
combine it in a quick prompt hopefully more precise than what I actually put up here with other documents maybe your
1:39:48
road map or your team's project database and you can directly have your meeting notes interact with all the other
1:39:54
artifacts you put on notion and with notion agent able to take action in databases it can then go do the updates
1:39:59
for you. So I now get out of my team meeting and within 10 minutes everything we talked about is appropriately
1:40:05
reflected back in the spaces where we live in notion. The last one, find a way
1:40:10
to do AI dictation voice to text. Uh it's a little different than meeting notes. It's literally put your cursor
1:40:16
anywhere on your desktop and hit a button and you start talking and it starts to record everything you you say
1:40:22
and then it edits it to be even better than what you said. Uh there are tools like Whisper Flow, Super Whisper, Mac
1:40:29
Whisper. They all have the word whisper in them. That's how you can find them. Pick one uh that you like. But this is a
1:40:35
great way to again be able to really speed up that precision journey. I have support team members who can now voice
1:40:41
type, we call it 200 words per minute, uh which is twice what they could before. And so if we go back to this
1:40:47
idea of faster and better, that speed is what drives their ability to now put better answers out to our customers, get
1:40:53
through more work, think a little bit more deeply. So once you have your team good at prompting, the next big phase is
1:41:00
knowledge. And for knowledge, I think this is a misunderstood space still in
1:41:05
AI. Uh it's a place where, and you'll see in a bit, I don't always agree with notion on how to approach this. Um but I
1:41:11
do know one thing, and that is that knowledge managers have gotten a bad rap. Uh, I think it is often seen as an
1:41:17
esoteric kind of dorky field. Knowledge management work is often on the bottom of people's list of things they have to
1:41:23
do in any given week or month. And I think that the AI era is going to fundamentally remake and re-elevate the
1:41:31
importance of knowledge management in companies. And frankly, if you're earlier in your career and you want to find a place I think will be useful for
1:41:37
another generation, knowledge management. And part of that is because if you've ever worked at a big company,
1:41:43
you probably lived in this, which is you have a dozen different tools where your knowledge is stored. It is routing the
1:41:49
information from that to a bunch of different places. It is maintained by half a dozen teams. Those tools have the
1:41:55
same information but represented in different ways. They're constantly out of alignment. They're vague. They're not kept up. And now we pour gasoline on the
1:42:03
fire, which is tons of AI that is asking those tools questions and then probably telling your customers information that
1:42:09
it found. And you as the person or me I guess as the person who has to own the knowledge of the company, I'm pretty
1:42:15
terrified that I'm going to lead to a lot of meh customers from the setup. And so we've really this year doubled down,
1:42:21
triple down on knowledge management. It's a first class problem at RAMP. Uh and we've figured out what we think A+
1:42:27
AI knowledge means. It's really four things. It's being single sourced, maniacally accurate, meaning keeping it up to date. Crystal clear, describing
1:42:34
everything as clearly as you can. And then you have to teach the team to actually use it wisely. Let's go quickly
1:42:40
through each of those. First one's kind of the notion sales pitch all over again. Right? When we deployed notion at
1:42:45
the beginning of the year, in addition to the AI functionality, we loved that it let us consolidate half a dozen tools
1:42:50
into one. And it wasn't just because of the cost savings. It's obviously nice. helps me like get people to agree to
1:42:55
roll out notion. But the real benefit is now you have a single place where all of the knowledge that is important about
1:43:01
your product and your customers and your policies lives and it can interact with each other. Now when something on the
1:43:08
product roadmap rolls out, it can automatically update our SOPs about that part of the product. When a new set of
1:43:13
tasks is completed by a team, it can automatically reshape the wiki for that team if that needs to happen. So once
1:43:20
it's all in one place, it's much easier to decide what is the core source of truth for each thing that your company
1:43:25
does. And it goes further. We also take full advantage of Notion's AI connectors. So we've plugged Notion into
1:43:32
Slack, GitHub, Google Drive, Linear. I'm really excited about the MCP stuff. We hope to have RAMP be on this list
1:43:37
someday for all all of the Ramp customers in the audience. But the ability to use notion AI in addition
1:43:43
across notion across everything else in your ecosystem where information is stored that's a real crucial thing and
1:43:51
it gets even uh better. We've taken some more uh creative means to get information out of uh customer sources
1:43:57
of truth and CRM like Salesforce and Zenesk surveys from Sprig and so on. For those we often have really important
1:44:04
information but it's sitting in this silo that's really hard to break into. So we scrub it of personal information.
1:44:09
We have an amazing info security team that helps with that. Uh, and then we plug it through connectors into notion
1:44:14
as well. So now we have in notion a database of all of our sales feedback. That's really useful to combine with all
1:44:19
the information that's already there. So get everything into a single source of truth and then if you know me at all by
1:44:25
now, do feedback loops with it. Uh, feedback loops are great in the AI era. And so we have something called knowledge feedback loops. With your
1:44:32
single source of truth, you're going to be using it all over the place. As you use it, it's going to fail you. as it fails you, you can have people uh submit
1:44:39
those failures. You can have AI submit those failures. You can have AI monitors that are built on notion custom agents
1:44:45
just looking for places where it clearly didn't do the job. AI can also research and propose the updates based on those
1:44:51
signals and then all your human is there for is to serve as the core arbiter of what is actually true and that allows
1:44:59
your knowledge to get better and better and be used in more and more places. Let's bring this to life though quickly. So this is our SOP hub. This is where we
1:45:06
store all of the policy documents for ramp. If I open my notion AI agent, I can ask it a question about, you know,
1:45:12
PDM expenses. Maybe I have a customer that wants to know about uh a specific policy. All right, notion ponders. It
1:45:19
does its work and yes, we get the right answer and beautifully we get to click right into the article that that answer
1:45:25
was sourced from so that we can confirm that it's actually true. Make sure we're not hallucinating. But let's say that I
1:45:30
know that this is actually not correct. I was just in a meeting and it turns out there's now some exception to this general policy and if this gets sent to
1:45:37
the wrong customer, it's actually going to confuse them quite a bit. I can now tag in through notion agent a SOP
1:45:43
feedback database and I can just explain what I think is wrong and now notion agent who can modify databases goes
1:45:49
ahead and fully adds the row to that feedback database. This in turn creates
1:45:55
a signal to our knowledge manager who owns this piece of knowledge that something might be wrong. So that's
1:46:00
great. We've moved much faster. This probably wouldn't have been submitted at all, frankly, if it hadn't been so easy.
1:46:06
And they can take a look. They can figure it out. But it gets better. We've built a prototype of a custom agent on
1:46:12
top that is going to attempt to rewrite the article for us so that our knowledge manager doesn't have to waste time
1:46:17
actually figuring out what the intended fix is. We put in the triggers, we put in the instructions, we put in the tools
1:46:24
and access that it has access to, all the things that we just saw in the keynote. And once it's done, it checks a box just
1:46:30
to let us know it did it. And it pastes the new article down in that feedback. Now all I have to do if I'm the
1:46:35
knowledge manager is read through that, make sure that I agree this is actually a new part of our knowledge and put it
1:46:40
back in the original document. So a whole thing that would have taken probably a week sitting in a ticket queue may not have gotten submitted at
1:46:46
all is now easy. And so once you get good at your knowledge feedback loops, you can focus
1:46:52
on your data too. And here the real key is that most data teams are not notoriously super clear in how they
1:46:58
describe the data tables that they produce. Uh, and we've tackled that
1:47:04
head-on. Our data teams actually spent a lot of time being more knowledge managers writing these natural language
1:47:09
documents that really describe all the nuances and intricacies of how the data works. Stuff that was previously tribal
1:47:15
knowledge. And when it's done, you can build things like ramp research. This is a tool that we launched a couple months ago that literally you ask any question
1:47:22
in Slack, it'll think for a couple minutes. It'll use those descriptions of our data. It'll produce a full report uh
1:47:28
along with all of the data queries that it used to generate that report that you can then use to go and and do other work
1:47:34
with that question. The beauty here is it's the most clear, faster, and better story I can tell you. The blue line here
1:47:40
is how many manual uh data analyst requests we were getting before we launched Aramp Research. The green is
1:47:47
when we launched the tool. We now have 15 times more people using the data of ramp to do their work. And it's because
1:47:53
it's so fast and easy to do. So faster drives better drives faster drives better. And finally, the one place that
1:48:00
I disagree with notion's design philosophy today is that the default when you open knowledge as a user is all
1:48:06
sources I can access. I do not want my team to use all sources I can access all the time. There's a lot of bad stuff out
1:48:12
in our Slack. There's a lot of years old stuff out in Google Drive. I maintain places that are sources of truth. And
1:48:19
so, one of the nice things that notion does do is it lets me pick out the specific places that I want to access
1:48:25
knowledge. But it's really the onus being on the user of that knowledge, the person who wants something to be a
1:48:30
little more specific. And I think all of you who are writing prompts should also be thinking about the precision of what knowledge scope you actually want to be
1:48:36
looking over so you don't get yourself into a situation of outdated info. And so that brings us to the last piece,
1:48:43
workflows. Uh we want everybody to be able to build their own way out of their own workflows. And so this is really
1:48:50
about scaling doing things one time to a thousand times really easily. And the challenge that this has historically
1:48:56
posed is, yeah, it's easy to write a great prompt. Not easy, but you can do it yourself. It's something you can do
1:49:02
yourself to put the right knowledge and information into that prompt. It's still very technical feeling. Requires
1:49:07
engineers to go and like hook that up. So every time Slack comes in or every time something fires in Salesforce,
1:49:13
something happens. And so when we talk about workflow, it's really about no longer needing to rely on technical
1:49:19
teams to do all those connections. And in the last few months only have the tools become good enough that we finally
1:49:24
feel good that we can put workflows in everyone's hands. We use a tool called Gum Loop. There are others like it, but
1:49:30
we find it to be really intuitive to emblematically say what the workflow concept is. You take any input and
1:49:36
trigger for work. You pull in all the information you need. You start it. You do AI with it, the prompting and the
1:49:43
knowledge stuff. And then based on what you get, you put the output somewhere else.
1:49:48
The way that Gumlu makes it seem so straightforward is that it gives you all of the nodes for the inputs and outputs
1:49:54
separately and then you can spend all of your time and craft focused on the AI in the middle. And so we've seen multiple
1:50:00
teams get value out of this. Our product team uses it to do weekly updates for their projects in Slack. It
1:50:07
automatically summarizes from four or five different sources and puts it all in one place. Our sales team is getting
1:50:12
pretty good at getting all of the deep web research that they need done uh and automatically drafting emails into their
1:50:19
outbox to send to prospects. And our operations team builds more complex flows that will answer questions when it
1:50:24
can and uh branch off and ask experts for answers when we don't know the answer.
1:50:30
I think what you just saw with custom agents is handinand glove to this as well. I think one of the things that notion does especially well is it can
1:50:38
use a lot more intuition inside of the walls of notion. And so yeah, the sales feedback database that we use, for
1:50:45
instance, we now have a custom agent whose job is I need to come up with a better name and a hat, I think, but
1:50:50
sales feedback categorizer, it takes the sales feedback in, tries to find something in our road map that would
1:50:57
address that feedback, and if it's say a feature that's already in a private alpha or something, we'll go ahead and
1:51:02
just draft the email for the account manager to be able to potentially oneclick enable the person on that on
1:51:08
that tool uh solve their issue and move them forward. So, I think there's a lot more work we're still going to do over
1:51:13
the rest of the year on workflows, but the pieces are now in place. And so, that's really the system. Uh,
1:51:19
you have your prompt, what precisely do you want? You have your knowledge, what do you need to know to do it? And you
1:51:25
have your workflow, which is when and where you want it. And so, the pattern becomes clearer and
1:51:30
clearer. We've spent this year really focused on moving fast and executing here because at the end of the day,
1:51:36
again, we want to build a company that is not just button pushers. We don't want to buy into kind of the myth of
1:51:42
doomers and zoomers and boomers that you have to wait until it's easier for you. We want all of the people at RAMP to
1:51:48
have it be harder at first to go practice to go get good at this work because on the other side we believe not
1:51:54
only that it can make us the most productive company in the world, but we also believe that it can make work more fulfilling for the people at RAMP. uh
1:52:01
you get to really control your own destiny when you're the one who wakes up, decides what you don't like about your job and feels enabled on the tools
1:52:08
to go and build your way out of it. It also compounds, innovation compounds in this space. So the team members who are
1:52:15
out front and are getting the hang of things, they work faster and then they can share what they they did and other
1:52:22
people can adopt it and they might not adopt a 100% of it. They might take the draft and they might tweak a few things to their liking. they might swap out the
1:52:29
notion hat and put in some new instructions. And then the last piece is it's fun. Like I've had more fun this
1:52:35
year building random stuff with AI, even when it fails, than I did when my job
1:52:40
was more of a button pusher job. And so I think it's something where if you take the time to get yourself really
1:52:46
productive, you get two dopamine hits along the way. One is when you crack the problem of how to automate the work that
1:52:52
you didn't like doing. And the second is when on the other side it goes quiet for a minute because you don't have to do
1:52:57
that work anymore. You have more time in your day. And so as a company that's geared towards saving time and money, it
1:53:02
feels as internally consistent as it can be to be taking this approach with RAMP.
1:53:07
So closing out, what are you going to build? Uh you've been here for an hour and a half now. You've heard about a ton of tools. You presumably all use Notion.
1:53:15
If you're not yet full users of Notion AI, like pay them their money and get the full suite of tools. It's a great
1:53:21
building place uh to learn this stuff, but get down the learning curve. I think a lot of people uh that I talked to
1:53:27
still seem to be waiting for something. It's not going to come. Now's the time. And so, if you can get good at the prompt, if you can get good at
1:53:33
understanding knowledge, if you can learn how to put it into your own workflows, uh what magical thing that
1:53:38
felt out of reach a couple hours ago now feels achievable? Um what are you going to build? So, that's that's kind of what
1:53:45
I wanted to talk you through. Thank you. If you want to learn a little more about RAMP, uh the QR code has some info on our team and products. But really
1:53:52
excited for the presentations today and I appreciate it.
1:55:08
Heat.
1:55:28
Heat.
1:55:47
Hey,
1:56:01
perfect.
1:57:02
Hey, hey, hey.
1:57:51
Heat. Heat.
1:57:58
Heat.
1:58:10
Hey, Heat.
1:58:24
Heat. Heat.
2:03:04
find.
2:12:24
Heat. Heat.
2:15:39
Heat.
2:15:59
Hey. Hey. Hey.
2:17:46
D
2:18:10
hey
2:20:06
Hey, hey, hey.
2:23:03
Heat. Heat.
2:23:51
Hello.
2:23:58
Who's
2:24:14
Doo doo doo doo doo.
2:25:37
Ding.
2:25:44
Ding.
2:29:17
Hey.
2:29:31
Hey. Hey.
2:29:45
Come on.
2:29:55
Hey. Hey.
2:30:07
Hello.
2:30:22
Fall home.
2:41:59
Everybody's
2:50:56
Hey,
2:51:02
D.
2:56:32
Heat.
2:56:44
Heat.
2:56:53
Heat. Heat.
3:00:58
I love that.
3:03:36
Hey, hey, hey.
3:04:07
Heat.
3:04:26
Heat.
3:09:00
Please welcome Dave Desperetes, Alejandra Sien Vueos, and Jenny
3:09:06
Familano.
3:09:14
Hello. Hello. How's everybody's lunch? Good. Great.
3:09:19
Well, nice to meet you all. My name is David Despus. I run a notion consultancy. I'm a Notion solutions
3:09:26
partner. uh run a company called Workcraft Labs in Miami and mostly work
3:09:31
with venture capital firms figuring out workflows and solutions to work smarter
3:09:37
and move faster. And I'm here with two other amazing
3:09:42
Notion ambassadors and consultants to share our stories of not only how we
3:09:48
joined the community, but how we built our entire businesses around notion. So
3:09:54
hopefully you leave inspired but also with some concrete ideas on how to work
3:09:59
smarter with notion especially with all of the things covered this morning.
3:10:05
A quick show of hands. How many of us have felt overwhelmed with new tools, tech, AI?
3:10:13
Got it. Okay. Myself included. I can say what I help clients with has changed
3:10:18
dramatically. Client projects went from building straightforward Notion workspaces to
3:10:25
redesigning entire workflows, usually with a combination of Notion automation
3:10:31
and AI. When clients reach out, they're usually they've usually seen other teams build
3:10:39
these powerful workflows. They're excited about the possibilities, but are also overwhelmed, asking questions like,
3:10:47
"What tools do we use? What do we do in notion? What workflows do we keep?
3:10:53
Everyone wants the outcomes without the complexity part.
3:10:58
We have access to a seemingly infinite amount of information. Our tools are
3:11:04
smarter than ever. And we're more connected than ever. And this kind of made me think this feels a lot like the
3:11:12
Renaissance. I know Ivan went back to 1900. I'm going going even further back.
3:11:17
When we unpack why the the Renaissance was so transformative, it really came down to three components.
3:11:23
First off, guilds became a thing. They brought people together, often from different trades or backgrounds, and
3:11:31
people now had physical spaces to share ideas, collaborate, and even train each
3:11:37
other on specific skills. When we think of Brunoleski, the architect of the famous Florence
3:11:43
Cathedral, he was actually trained as a goldsmith, but being in a guild meant he had access to craftsmen, artisans,
3:11:50
engineers eager to share their expertise. Guilds were spaces to share knowledge
3:11:57
and we have our own spaces today. One of which completely changed my professional
3:12:03
path. Now around the same time, the creation of the printing press meant books that
3:12:09
were once limited to libraries or elite institutions were now available at
3:12:14
scale. Authors now had channels to share their work, to share their writings. And
3:12:21
this allowed new ideas to spread and discourse to really thrive. And lastly,
3:12:27
technological in innovation made structures like Brunoleski's dome possible. Having been under construction
3:12:35
for over a century, no one knew exactly how to finish it. It was only through
3:12:41
coupling new techniques with classical practices from ancient Rome that it finally got built.
3:12:49
The combination of these three things, community, knowledge, and systems
3:12:55
evolving at the same time, led to a total shift in how people learned, but
3:13:00
more importantly, what they built. So, what does this have to do with today? We'll get into it. In many ways,
3:13:07
my experience with Notion probably runs parallel to many of yours. I'm going to rewind back to early 2020 teaching high
3:13:14
school art and design in Brooklyn. Any New Yorkers out there? Got a few. All right. Long flight. Uh, from one day to
3:13:23
the next, our school, our day-to-day routine completely came to a halt. No
3:13:29
more clear schedule, no more routine. my wife, daughter, and I figuring out the logistics of our days at home.
3:13:36
I felt like I needed a way to structure my work schedule and dad duties and time
3:13:41
blocking in Notion was my way of doing that. I was fascinated by the idea that you could build tools to fit your needs,
3:13:49
not the other way around. Thankfully, there were already creators like Marie Puland and August Bradley,
3:13:56
which many of us know, generously sharing not only what they built in Notion, but exactly how they built it.
3:14:04
So, as I started building my own tools in Notion, I decided to do the same. Just share what I was building with the
3:14:10
hope that it might inspire others to build their own custom workspaces. And I always cringe looking at this video
3:14:16
because it's now 5 years ago and the mic was terrible, the video was terrible, but you live and you learn. Fast forward
3:14:23
a few years and I was tracking literally everything in Notion from journal entries to workouts to habits. Again,
3:14:31
I'd never seen any other tool be able to do this. And the more systems I built
3:14:36
for myself, the more I wanted to build. And as I shared more about my notion setup, I eventually started selling
3:14:42
templates. people started reaching out asking if I could help them build their workspaces.
3:14:49
I was a little unsure about this. So I went to the interwebs and I looked for other notion consultants and thankfully
3:14:56
Francis Oderea Matthews in the UK was hosting Twitter spaces uh weekly or
3:15:01
bi-weekly talking about all things Notion and usually I would chime in and pick her brain on what it was like
3:15:07
working with clients on notion projects like was there enough demand? Were there
3:15:12
any red flags? Um, what were some things to look out for?
3:15:17
Eventually, I jumped into notion consulting, but it was only because I had access to a global community of
3:15:24
people learning, sharing, and building in public. And fast forward to today,
3:15:29
the notion ambassador Slack workspace is filled with tons of channels on everything from writing formulas to
3:15:37
making YouTube videos to selling templates. Our Notion Miami community is watching. Hope you all are having a
3:15:42
great time. And the three of us are all working with businesses, all of which
3:15:48
run their entire operation in Notion. I literally would not be here if it
3:15:53
weren't for other Notion creators generously being willing to answer questions, jump on calls, meet in
3:15:59
person, and share what they knew. As technology keeps accelerating,
3:16:06
finding your communities is essential in today's productivity renaissance.
3:16:11
Having a network, circles of people sharing, collaborating, supporting is, I
3:16:18
believe, your greatest competitive advantage. But this doesn't change the fact that
3:16:24
our tools are getting better and smarter every day. And that could be overwhelming, as many of us know. I can
3:16:31
tell you personally, it feels like I'm playing catch-up. Like with all the Notion AI stuff, I just discovered that
3:16:37
you could move comments to the bottom of the page in a database page. Like, who knew you could do that? I don't know.
3:16:42
It's news to me. Maybe you guys are already doing it. Um, but Notion today looks similar to what we
3:16:49
knew when we started using it, but it's so much more powerful. Everything in our workspace is available with just a
3:16:56
simple prompt. And as you all have seen, Notion AI is taking on a more proactive
3:17:01
role in our workspaces, taking on a lot of those manual, repetitive tasks.
3:17:07
Building personalized workspaces is easier than ever, but building workflows
3:17:13
for teams is entirely different. Some team members are at the bleeding edge of
3:17:19
new tools and tech. Other team members still like to use Post-it notes, which
3:17:24
that's okay. I'm not I'm not judging. And there's everyone in between. It's less about what you build and all about
3:17:31
aligning on how the changes you're making is helping your team work smarter.
3:17:37
When you could build anything, the key is deciding what to build and understand
3:17:42
why that matters. I'm going to share a simple framework called Labs that has been helpful
3:17:49
getting teams to go from ideas to action. making sure we don't spend too much time
3:17:56
spinning our wheels optim optimizing systems that just don't work. This
3:18:01
reminds me again of Brunoleski where we're using triedand-true methods but in an entirely new context.
3:18:10
So first before building anything we have to understand as a team or as a solo person uh or solarreneur what's not
3:18:19
working using notion forms for example any notion forms users
3:18:24
here and there the rest of you need to get on it fast so uh using notion forms we could easily
3:18:31
gather info on what's taking up too much time what's very manual what's disorganized just getting all those
3:18:38
ideas in the same I'll say one area that comes up quite a bit probably the most
3:18:49
responses like our meeting agendas are very well organized but they then go off
3:18:54
the rails or we talk too much about plans and not much on execution. Uh
3:18:59
these things kind of come up over and over but we can only decide what to focus on once we've put everyone's ideas
3:19:05
together. And the key here is deciding on what that one thing is and
3:19:11
architecting a solution to only that one thing.
3:19:16
Even with just a few bullet points, you'd be surprised how a simple AI block can recommend what to build and why.
3:19:24
Sorry, last show of hands. AI blocks. Who uses them? Okay, I know a lot of us are talking about the AI sidebar and our
3:19:31
agents, which are fantastic by the way, but AI blocks are kind of like a sleeper feature where you can just slot them
3:19:37
into pages and into database templates. Definitely check them out if you haven't used them. But this is about
3:19:43
architecting a solution, defining what needs to be built, how it should work, and most importantly, how do we know if
3:19:49
it's effective? From here, we can get building. This is
3:19:55
also an area that's gotten incredibly easy with Notion's database builder.
3:20:00
When this came out, I was wondering if I'd be out of a job because it was so good. But thankfully, there's a lot to
3:20:05
do, so we're good there. But if you're a product manager, you've probably diagrammed out workflows or sequences.
3:20:12
And this same practice is great for teams. Again, a simple prompt can turn your raw
3:20:18
notes into a fully functional, in this case, mermaid diagram. Uh that is super
3:20:24
useful just to understand what are the connections between databases, what's being added where, um what do we
3:20:30
reference versus what do we add? Uh details like that. With the database builder, all you need
3:20:36
is plain text to get a fully functional database. Customize with the exact properties that you need. Now, with
3:20:44
everything AI related, we always want to meticulously review what's being
3:20:49
suggested, but even that first step saves tons of time. We don't have to build from scratch
3:20:55
anymore and we don't have to tweak an existing template to try to make it work
3:21:01
for our team when it's just not built for us. Once we've built new spaces, use them,
3:21:08
refine them, the most important part of the process is making sure our team is
3:21:14
not only aware of the changes, but aligned and bought into how these changes are going to help us all work
3:21:20
smarter. This is often the most overlooked step, but is the most critical.
3:21:29
By focusing on one pain point at a time, we can make changes that free up hundreds of hours. For example, as a
3:21:37
solo consultant, tracking discovery calls in with Notion AI's meeting notes
3:21:42
and having a prompt just copy and pasted into Notion AI, which is now not even
3:21:48
doable with custom agents. That's kind of saves a couple more steps. Uh can create a proposal draft that is not only
3:21:54
fast and formatted in the way I want, but is using my tone of voice and in a lot of cases actually brings up details
3:22:01
from conversations that I might have missed. And for one client who migrated their entire investment team to Notion,
3:22:08
meeting pages now surface every relevant update. Meeting pages that have
3:22:13
fundraising updates, prospective potential investments highlighting which projects or initiatives are blocked and
3:22:19
need some attention. When you can build anything, the real value comes in the decision of what to
3:22:27
build. Focusing on one problem at a time allows us to identify the root of the
3:22:32
problem, brainstorm approaches, and build our own solutions.
3:22:38
This simple exercise has been super useful for solarreneurs to tiny teams to
3:22:45
venture firms managing over a billion dollars in assets with just a few simple questions to get the conversation going.
3:22:52
It's less of a form to fill out and more of a prototype to start working with Notion AI and help make those decisions
3:23:00
together. So, if you have some ideas that are top of mind for your team, grab
3:23:05
that, you know, guide using the QR code and get building.
3:23:11
In our productivity renaissance, we have a massive opportunity to completely
3:23:17
rethink how we work. we can focus more deeply on the ideas and questions that
3:23:23
matter most. So, as we continue to navigate this
3:23:30
renaissance, I'll leave you with a simple question. with our own communities in the form of in-person and
3:23:38
online groups with a wealth of knowledge at our fingertips and with tools like
3:23:44
notion to be able to build our own personalized systems to work smarter.
3:23:49
I would love to know what will you build next. Thank you so much for having me
3:23:55
and please welcome Alejandra Sen Fueos to the stage.
3:24:04
Thank you so much Dave. Today our notion community is global but
3:24:10
regional guilds like the Spanish one were a very important part in shaping the community we have now. Hi everyone,
3:24:19
my name is Alejandra. I'm a certified consultant and ambassador and I'm really happy to share my story with you today.
3:24:27
My journey looks a lot like this. And as you can see, these are the six years
3:24:32
that changed everything for me. All began in 2020 while I was looking
3:24:38
for a place to document a trip I took with my partner Daniel. We wanted a
3:24:43
place to document all the wonderful experiences that we had on that trip.
3:24:49
And that's how we found notion. And this is the very first notion page that I
3:24:55
made a travel journal. And while creating this travel journal,
3:25:01
something clicked. Notion could be anything. A travel journal, a note, a
3:25:07
note taker, a meal tracker, anything. So I started thinking about all the use
3:25:14
cases I could use notion for. But at the time, I didn't have all the technical
3:25:20
knowledge. And for me the best way to learn is by practice. So I practiced. I
3:25:26
started creating templates. My partner Daniel also started creating templates. And together we built a huge template
3:25:34
library of 70 notion templates. We saw an opportunity to share our
3:25:40
knowledge and we put all of the templates on a website. And that's how we created notion. And this is us in the
3:25:48
public library brainstorming on this project. We also created a social media accounts
3:25:54
to share our work and people began to notice and to reach out. And this tweet
3:26:00
is very special to me because you can see Philillip asking for a notion template and me sharing that very first
3:26:07
template that I did, the travel journal. And the best part of this interaction is
3:26:14
that Philillip also became a notion ambassador and we were really happy that
3:26:19
we could also meet him in person a couple of years ago. And then in 2021 we got the email. We
3:26:28
were official notion ambassadors.
3:26:33
We got access to the community and we discovered that other ambassadors
3:26:39
already knew about our work and also about us. And we also found there some
3:26:46
familiar faces from previous social media connections and we felt welcome
3:26:52
from the day one. and the community became a very important part of my
3:26:58
journey, especially when we decided to turn our
3:27:03
notion project into a more professional one. Thanks to the meetups, the Slack
3:27:09
conversations, the showcases, we got inspired and we went from onepage
3:27:15
templates to complete systems that solve real problems and that also generate
3:27:22
revenue for us. This template is special the business plan because this template generate us
3:27:30
our first dollar online and after building business oriental
3:27:38
templates I got experience and I was ready to go allin in ocean and I got
3:27:44
certified in 2022 and then came a very special milestone.
3:27:52
Spanish community was very active but notion was not in Spanish yet. But in
3:27:58
2023, Notion and Espanol launched and we celebrated with an event in Madrid. And
3:28:06
the event was the perfect excuse to meet more Spanish ambassadors and also part
3:28:11
of the team that was in charge of notion localization in Spanish.
3:28:20
And after the Spanish launch, more Spanish companies joined the notion wave
3:28:26
and our client base grew not only in Spain but also in Latin America. And
3:28:34
this became the foundation of our notion focused business not only for templates
3:28:40
but also for consultancy and education. So we move from notion to what is now
3:28:48
Noxen Studio operating internationally as you can see in this notion map. Be
3:28:54
sure you try that feature soon. And working with companies across
3:29:02
regions made me realize that cultural context matters. It's not the same to
3:29:08
build for Spanish company that for a French company or a USbased company.
3:29:15
And because cultural context matters, building in ocean should focus on people
3:29:21
and asking people the right questions. And this is something we try to do every
3:29:27
day in our business. And
3:29:34
this is something we implemented when we work with clients like Ole. This first
3:29:42
approach is key in our discovery goals and it was one of the reasons we secure
3:29:47
Ole as a client. O is a Spanish company in the tech industry and at the time
3:29:53
they were facing these four core notion challenges. And to make sure we ask the right
3:30:01
questions, we have this simple framework with only three components and three questions that open very important
3:30:08
discoveries for us and help us build better systems in ocean.
3:30:15
And for Ole, we start by asking what already works here.
3:30:22
A fluent team, strong routines and collaboration habits.
3:30:27
And then we look for the growth catalyst. We needed to make the metrics
3:30:33
visible and also connect the project management team with the sales team.
3:30:39
And finally we discover the limitations security requirements and also different
3:30:45
KPIs definitions across teams. And with these three elements aligned we could be
3:30:51
a system with confidence. Now let me show you what changed.
3:30:56
We build a system based on their assets especially their collaboration habits.
3:31:03
One share collaboration page for projects. clear action buttons for all
3:31:09
teams and key information visible to everyone in the team.
3:31:15
We also built in notion a real source of truth. The metrics space was very clear,
3:31:22
very simple instructions. Metric definitions were were agreed and aligned across teams and we also added
3:31:30
some visual charts so we can easily spot trends.
3:31:36
And the result full adoption operations team, finance team, project
3:31:43
management team, all tracking hours and tasks in Ocean.
3:31:49
We also achieve real data centralization. All teams were updating metrics in real
3:31:56
time and those metrics were accessible to everyone with align definitions too.
3:32:05
And this is my invitation for you today. Take this framework, these three
3:32:10
questions, and apply them to your workspace. It doesn't matter if it is a professional one or a personal one. I'm
3:32:17
sure you will discover something valuable. And finally, remember that your most
3:32:25
valuable asset is the notion community. So don't be afraid to connect, to ask,
3:32:32
and to share. Thank you so much. Please welcome Jenny to the stage.
3:32:39
Thank you.
3:32:45
Alejandra just reminded us that our community is our most valuable asset.
3:32:50
And I want to share exactly how my community helped me understand and discover something I never expected.
3:32:56
that the way I had been developing my own creative process could be the foundation of helping me build a
3:33:03
business to help other people organize theirs.
3:33:08
My name is Jenny Familiaro. I'm a designer and I'm a notion consultant. And for me, design thinking
3:33:16
isn't just about making things look beautiful. It's about solving problems
3:33:21
visually. That approach completely changed how I work and eventually how I help other
3:33:27
creative teams work too. And my journey with notion started when
3:33:32
I was working at a content marketing agency when I was tasked to add copy to
3:33:38
these graphics for Blizzard's game launch and something felt a little off
3:33:44
about it. the text for tomorrow because of the length of the word made
3:33:51
it so it was visually smaller. It didn't have the same impact that the other two
3:33:56
pieces of copy did. So I did what any designer would do.
3:34:02
I tried to make it work. I explored different variations, different layouts.
3:34:08
I assessed readability and I wanted to maintain the same system that I had established in the other two pieces of
3:34:14
graphics. So, I organized all of these options in notion, not to just look at my work that
3:34:23
I've been doing, but to help paint the picture for my art director, to tell him exactly what I had been struggling with.
3:34:29
And when he showed it, when I showed it to him, he knew exactly what we needed to do and effectively change the copy.
3:34:37
Woo! Yes, designers can change copy.
3:34:43
And that moment taught me something very important that visuals can help guide
3:34:49
action. And notion is the tool that helps me take action on my visual ideas. As a
3:34:57
designer, I was able to work on a variety of different projects from infographics to type design projects and
3:35:04
brand visual identities. I map out my whole process in notion to aid in review
3:35:09
for my teams as well as my clients and so that I could know exactly what steps
3:35:15
I needed to take next. And in my head I thought this is how
3:35:20
everybody worked until I was walking through one of my
3:35:26
motion graphic projects. We were on design round three. We were walking through all the expectations, all of the
3:35:32
feedback that had been implemented and just showcasing all of these different deliverables within the feedback
3:35:38
tracker. Everything was organized. Everything was going smoothly.
3:35:43
Someone from that call, one person reached out and said,
3:35:49
"Hey, you seem to really love what you do.
3:35:57
Can you show me how to do that? Yes.
3:36:03
And that's when I realized that showing my process could guide others to take action on theirs.
3:36:12
I started sharing with people so how I work within my notion workspace
3:36:17
with intention. So I'm not just decorating. I'm leaving breadcrumbs so that future Jenny can have an easier
3:36:25
time making decisions. So, my creative workspace became my
3:36:31
creative compass. Instead of staring at a blank page, I was able to look at the
3:36:38
visual structure and know exactly what steps I needed to take next. And I found that when I'm detail- oriented at the
3:36:43
very beginning of a project, I'm able to spend more time doing things that I absolutely love doing.
3:36:50
which is creating and making because using design thinking for my
3:36:57
notion isn't just surface level aesthetics. We spend so much time in our
3:37:02
notion workspaces that it should reflect what we value too.
3:37:07
And for me, I value clarity, connection, and creative freedom. My notion
3:37:13
workspace was set up by me for me and had to reflect my best work. And in
3:37:20
order to do that, the workspaces that I build and use for others should embody this approach, not just organize it.
3:37:29
So what does my full creative ecosystem look like today? Here's a real example. So this is
3:37:36
my road map template that I use as my go-to template for whenever I kick off a
3:37:42
project with a new client. This is so that we can show how process can help influence exactly what we
3:37:49
deliver and how are those things connected. I group based on process dashboards, databases and sources. And
3:37:57
this allows teams finally to see exactly how all of their digital assets, how
3:38:03
they all fit into the puzzle of the system that we're building together. using hierarchy and database
3:38:10
relationships to highlight the risks missing requirements in order to prevent
3:38:15
scope from changing. So no surprises for me, no surprises for
3:38:21
my clients. We also have a timeline view that shows
3:38:27
details of where we are in phases and progress to stakeholders. One client
3:38:32
actually said that it turned their task management from a chore into a celebration because now we're able to
3:38:38
see exactly where we were able to go, where we've come from, how far we've come along in the process.
3:38:45
Tracking small ideas and thoughts to bigger ideas just so you're making
3:38:50
decisions based on real data and not just emotional ones. You can utilize every little instance to help drive
3:38:57
action towards something that is meaningful to you. adding your visual brand so that way you you and your team
3:39:03
know exactly that this space is crafted for you and for them.
3:39:09
Utilizing colorcoded with color theory, it helps tell the story at a glance so
3:39:14
you know exactly what you need to do next.
3:39:20
And the secret that I've discovered is if you give people their creative space
3:39:25
to get messy, they'll be able to show up easier to maintain team transparency.
3:39:32
My personal dashboard philosophy is I need to be transparent about how I
3:39:37
actively contribute to these goals and also the creative freedom to find satisfaction.
3:39:44
This means that individual ownership plus collective clarity equals systems
3:39:50
people can take meaningful action on. And so what started as a journey from my
3:39:56
own solo designer dashboard, now I get to implement it for other small creative
3:40:01
teams. So I started putting myself and my business out there more at around 2024
3:40:09
when I became a notion ambassador and I also met Diva Alejandra
3:40:14
at Make with Notion last year. But during this time I got stuck in my
3:40:22
head about between being a designer this identity that I wanted to hold on to and
3:40:28
my new role as a consultant. I fell into this comparison mindset that I would see
3:40:34
wildly talented people putting out things like YouTube videos, templates, and making a killing out of it. They
3:40:41
were things that I wasn't doing. I was scared.
3:40:49
But business is full of scary new experiences.
3:40:55
And I figured that as long as I kept looking forward with my creativity and my knowledge as my tool and my weapon
3:41:02
into the great unknown, I could learn to share the mission and the business that I built.
3:41:10
So despite being really scared and freaked out at the last make with notion, I signed up to share my
3:41:16
presentation on mastering visual identity with my fellow ambassadors in this
3:41:22
community. And instead of competition, I was met with support. They became my
3:41:29
cheerleaders. They became my friends. And they also thought that complex
3:41:34
systems were worth celebrating and unpacking together. And everyone made me safe. They made me feel safe to be
3:41:41
present and capture the moments. And so I connected with ambassadors,
3:41:47
other consultants, Notion team members, and I just kind of sat on the sidelines
3:41:52
drawing. And it made me feel so comfortable and safe because I was able
3:41:58
to listen and learn from the people around me during all of these different conversations.
3:42:03
And it became a fun way to capture memories of people like Nick James who wanted a picture drawn of him. So I
3:42:09
happy happily obliged. And something that I learned going into my fourth year
3:42:14
of being a notion consultant is that building a business is full of unknowns.
3:42:20
That design didn't need to be this separate part of me
3:42:26
because everything that I can do right now can be a tool for how I remember and
3:42:32
connect with people. And that's how I got here. And it makes
3:42:38
me think about why all of this matters. What my work does for my clients. The
3:42:44
teams and individuals that eventually graduate from working with me because they're able to navigate their own systems without me. That there are
3:42:50
agencies winning bigger projects because they're able to manage and communicate their value and their process clearly.
3:42:58
And that there are creative professionals finally feeling valued for their work, getting the raises and the
3:43:04
promotions that they deserve. And what do I get?
3:43:11
I get full-time work that is 100% aligned with me. Proof that
3:43:19
structure fuels creativity when designed with intention.
3:43:25
My story told at Make with Notion and a business that can help support me and my
3:43:31
family because it was always about creative confidence and because when you can see
3:43:36
your process clearly, you can communicate your value. And when you can communicate your value,
3:43:43
opportunities will head your way. So now it's your turn. I'm going to
3:43:49
challenge you to look at your workspace with a designer's eyes. You're going to play art director for a quick second and
3:43:55
you're going to be playing looking at your notion workspace. What would you change first if you approach things visually? What are some pain points you
3:44:01
can get rid of? How could you design intention? Not just organization. What is the number one action that you want
3:44:08
people to take on that page and make it amazingly obvious to them? What workflow
3:44:14
or process can you make visible? Because you don't know that when you share how you do what you do, it's not just this
3:44:22
abstract idea anymore. It can be your next breakthrough, your next business on
3:44:28
what to build next. And this perfect idea that someone else
3:44:34
needed to know. Make it seen. Because with Notion's flexibility as a
3:44:40
creative platform, even the most unexpected ideas can find form, flow, and a future. when cultivated under the
3:44:47
right conditions. And your timeline can start right here, right now. You can find us over here at
3:44:54
the Notion community space right outside where you can connect with us and other community members.
3:45:02
Start building your story today. We'll see you guys there.
3:49:24
Please welcome Claire Vo.
3:49:38
Hello. Okay, I was warned. They gave me a warning about you guys in the back.
3:49:44
They were like, "You just had a big lunch. It's 2 o'clock. We might be over it." So, I've got to like capture
3:49:51
attention. The other thing is, unlike most talks I give, this is like a beautiful bright room. I can see all of
3:49:57
you. So, usually it's like the front row that's under pressure, but now I need energy from everybody else. um while I
3:50:05
talk through 30 35 minutes. We're gonna get in and out of here. You're gonna be so confident in the age of AI. It's what
3:50:10
we all want. Um so, thank you for being here. Thank you so much to Notion for having me. I love all the releases from
3:50:17
this morning. I'm super excited. Um but today I'm here to talk about AI and in
3:50:23
particular about how we can enter this age of AI mostly as organizations to be
3:50:29
able to ship confidently. I can ship a lot of stuff in the age of AI, but can I
3:50:36
feel good about it? So, we're going to figure out a couple tactics that I think are really useful, especially for teams
3:50:42
to really feel confident as you move into this new way of building and operating as individuals and as
3:50:48
businesses. So, this is me. I'm Claire. I'm the solar founder of Chat PRD. Somebody
3:50:54
backstage again was like, "How big's the team now?" And I was like, "This is this
3:50:59
is the team." Um, and so I'm building a AI platform for product managers. I host
3:51:05
a podcast called How I AI. I've been a CPTO many, many times in many, many
3:51:11
large organizations. And I've sort of done tech in every format you can. I
3:51:17
have raised venture capital and bootstrapped. I've been acquired. I've worked at big companies and publicly traded startups. I've been there. And I
3:51:25
have never seen anything like I've seen the last two or three years in our
3:51:30
industry. It's really transformed how I approach everything from a professional perspective. And I'm here to just share
3:51:37
some lessons that I think are really practical about how teams can address this transition themselves.
3:51:45
So the thing that I really reflect on at this moment is there is just a new
3:51:51
baseline. just like we're we're playing on new rules and everybody needs to
3:51:57
learn to adapt. And when people talk to me, they're like, "Oh, Claire, you just you have a different life than all of
3:52:03
us. You're like at home floating through a large language model. Your only friend is Devon. Like, you have nothing to
3:52:10
teach me about what it's like to be in a real company, be in a real team." But I've done this in large organizations
3:52:17
and I really think that you have to adapt and probably faster than even I
3:52:22
think the folks in this room that are probably more on the edge are adapting. And why I think you really need to adapt
3:52:29
is I'm a million years old approximately. My oh god my 8-year-old said to me the other day, "Hey mom, do
3:52:34
you know in the olden times like in the 90s?" And I was like, "Oh my god. Oh my god." Like the olden times. Like I did a
3:52:41
startup 10 literally 10 years ago like exactly 10 years ago and it took me
3:52:47
three months to build a prototype. I raised a million dollars when that was a big deal. Um my founder job was like
3:52:54
about acquiring customers and hiring the team and like talking to investors.
3:52:59
I used that money to like hire people and I spent all this time managing people and we like sat in an office and
3:53:06
I rented real estate. was very strange and I was constrained by headcount everything we wanted to do it was like
3:53:12
uh we can't do that because we don't have enough sales people or uh we can't do that because we only have this many engineers and I just felt really con
3:53:20
capital and people constrained and it was a really interesting experience
3:53:25
because now legitimately the prototype of chat pierd
3:53:30
was like 30 minutes um to build a v1 it's completely bootstrapped my job now
3:53:37
still about customers. They're still they're still the real boss. But it's about managing tools. I've set up I I
3:53:44
haven't spent my time hiring. I've sped up my time every time I think I'm constrained by a task. I think how do I
3:53:50
set up an automation to do this? And now I'm constrained by ideas. Like I
3:53:56
actually people are like aren't you going to hire engineers? I was like I don't have enough idea. like I'm I'm constrained by good user problem, you
3:54:05
know, margin of creative ideas more than I am about headcount or capital. It's just like a totally different world.
3:54:13
And so like the rules have fundamentally changed. And the question that notion asked me to come up here and answer is
3:54:19
like okay, the rules have changed. How are we supposed to feel good about our work? How do we feel confident? How do
3:54:25
we actually approach this new world when the rules have changed? And again, this shift is happening for
3:54:32
more than just startups. So like this is just an example. And I actually did
3:54:37
worked on Chat Purity for 18 months while being CPTTO. It was like totally a
3:54:42
side hustle. I like did like 5:00 a.m. to 7:00 a.m. and then like 6:00 p.m. to 8:00 p.m. every day all weekends because
3:54:51
I really felt like this investment in figuring out this AI native way to of working was one of the most practical
3:54:58
professional development exercises I could do as a technology executive. Like
3:55:03
I didn't think that I was going to be able to effectively do my job in the next three years if I didn't see what
3:55:09
was coming and what was happening in the house. for example, in our large team
3:55:15
established situation was just like 100% not what was coming for us. It wasn't going to be how our competitors were
3:55:21
going to operate. It wasn't going to be how new products are going to be built. And so I felt like it was imperative
3:55:27
that I learned this model myself and got my hands on it because I thought it
3:55:32
would make me a better leader and it did. The problem is like I whoopsied into an enterprise software business and
3:55:38
eventually had to go full-time. That's another story for another day. But really the intention of this initiative
3:55:43
was for me to figure out inside a team who is they going to compete with, how are they going to operate, how do we
3:55:49
need to operate, and like how can we get there faster and so I really do believe teams of all sizes need to figure out
3:55:56
their AI native operating model. And it's probably more extreme than what you're doing today.
3:56:04
So you practically what do you need in order to build this AI native operating
3:56:09
model? think you need like a good I know how McKenzie does presentations three things right okay we're going to do
3:56:15
three things you need an operating model and I really believe in systems and you'll hear me say this a lot in this
3:56:21
talk you need a new way of working you don't need like AI tools and you don't
3:56:27
need like tricks you need to think like how do we operate end to end how is it going to be different and then work
3:56:34
backwards into into an operating model second thing you need is culture and
3:56:39
Actually, this is truly the hardest part of this transformation. There is so much
3:56:46
entrenched in culture inside teams, small and large, that it is very, very
3:56:52
hard to change the way you do work if you do not have the right cultural
3:56:58
attributes. And it's really hard to build new cultural attributes inside
3:57:03
teams. And so this is a lot of what I was trying to figure out and continue to try to help people figure out is like
3:57:09
what are the right cultural tenants that you need inside organizations and how do you seed those so they help help drive
3:57:17
your transformation forward. Then you need guardrails like I am the poster child for like yolo AI. I tell
3:57:25
everybody like all the things that you're worried about with AI are like rich people problems like oh your vibecoded stuff doesn't scale. Oh good
3:57:32
for you. you have too many users. Like I'm on the far edge and yet like there be dragons. We're going to we're going
3:57:38
to figure out the things that we need to put on guardrails because again guardrails make people feel confident.
3:57:43
They're going to make your CEO feel confident. They're going to make your team feel confident, make your customers feel confident. They'll make legal feel
3:57:50
confident. So you you definitely need guardrails. So I'm going to go through talk through a couple things that I
3:57:56
think are really practical here that you can do that I think will scaffold you up for a successful roll out. in the
3:58:02
future. Okay. So, operating model and again I'm going to like stay with very like practical things that I think
3:58:09
everybody should be thinking about doing or practical things that I have approached that I think are useful.
3:58:15
So, the first thing is build the toolkit and define your flows. And I think a lot
3:58:21
of people uh and a lot of teams ad hoc start with number one which is like what are the tools we should learn like we
3:58:27
should learn to prototype we should learn all these great new notion features we should make sure we have a contract for this and can I legally
3:58:34
record everybody without telling them like they want tools and what they haven't done is taken those tools taken
3:58:42
the people on their team step back and said okay like what are
3:58:48
our repeatable processes What are the repeatable playbooks that we run every
3:58:53
day, every week, every month where we know these tools slot in and like how do we change our work around those those
3:59:00
flows? And so I have two flows and again I define them as like these are flows I
3:59:06
use. These aren't tools I use. They're not tricks I use. They're not like products I build. They're like truly like systems that I rely on that I know
3:59:14
everybody on the team knows about and that we like replay over and over and over. So I call this one the always on
3:59:21
team flow. It's like the bonus team flow. It's like how can we have
3:59:26
one set of systems that does a constrained set of work for us in a regular process over time. And this is
3:59:34
one that I use and you'll look at this and what you'll notice is it scares people. Sorry. Sorry I scared you. We
3:59:41
have we have roles up here like what role does this tool play? But I think it's actually a really useful mental
3:59:47
model to say if this were a teammate, what title would they have? If this were
3:59:53
a person on our team, if this were headcount I was asking for, what's the name of the headcount I would ask for?
3:59:58
Because it's the right, we're all in organizations, right? We all understand organization dynamics. It's a very
4:00:05
natural language and I think it needs to be extended to how we use these workflows. So, okay, PM define the
4:00:10
product and chat pod. Then I click a magic button and I get a prototype. Why
4:00:15
do I get a prototype? Because I think it's easier to communicate what I want with prototype.
4:00:21
Then um we pull those into cursor via MCP and we code the feature. So this is
4:00:26
like were I to don't tell I am not allowed to say replace. If I were to
4:00:31
augment the team with a new triad and that triad was mostly tools, what would
4:00:37
those tools look like and how would we feed those tools step by step? And you all do this all the time like you have a
4:00:43
product manager and they do this and then you have a designer and they do that like it is the flow but I find that
4:00:48
so many organizations don't just write down like this is the flow now and for this set of work you like just like you
4:00:55
would say you go to this PM you say you go to this flow and we get that done but then there's more to that I think where
4:01:03
people um miss out on AI right now in terms of defining their flow is they top
4:01:10
in maybe this like first chunk which is a very functional view like this is like the EP view product to product something
4:01:18
design something engineering code something then like you deploy it and you get something done but that's not
4:01:23
where work ends in your organization so the challenge I give to you is like pull that flow as far as you can on the
4:01:31
actual process that you're trying to do don't stay in your function or in your role which I see so many people doing
4:01:37
how can I as a designer use AI to design better, how can I as an engineer, how can I as a marketer and instead of
4:01:43
saying like how do we get stuff done in the org and then how far can I pull that
4:01:48
workflow through end to end and so we deploy a feature but then what
4:01:54
we do is like we write documents for it automatically and then this is my favorite one which is then I would send
4:01:59
it to product marketing and be like can you please get me users and we have an automated flow to get it into marketing
4:02:05
again like we could probably pull this even f we actually do pull it further Uh I should have put more side here. Uh
4:02:10
these docs go into support documentation. Support gets fed into an AI agent. AI agent triages requests.
4:02:18
Those requests actually go back and get prioritized. And again like that's how our functions work. I've just done the
4:02:25
effort to actually explicitly say this is a flow we do. This is a workflow we
4:02:30
do. And I think you could all take a step back and probably define four or five of these inside your company that
4:02:37
would really work for you. Another one is just my 247 uh intern flow. I run a podcast so I don't have an
4:02:44
admin for this podcast. So a lot of automations. I'm like h if I had like the ideal intern for the podcast, what
4:02:50
would they do? And like they would schedule a guest, which they do. We
4:02:56
populate notion um automatically. Prep emails get populated automatically.
4:03:03
Research happens automatically. Then we send guests emails um ahead of the show. I record it and
4:03:10
then we have a whole automated flow that again markets that that podcast episode.
4:03:15
And again I'm I'm giving this not as an example. Maybe this is useful to you. you're on a podcast, but more I think of
4:03:21
it as like if I had organizational structures that could support me, how
4:03:27
would I design those, what would they look like? And then what tools and people would I put in that process and
4:03:32
then publish it and say, I think this is a new way we can do work. And what I would say is a lot of these are analogous to how existing organization
4:03:40
structures, but you may find that you want to do something totally different. And I think that's a really good
4:03:46
exercise for folks to do. Okay, so first thing, define your toolkits, define your flows. You'll all
4:03:52
go and like whiteboard this. The second thing is, and I still think I still
4:03:58
underdo this. I do not dedicate enough time to automating the mundane. And so
4:04:04
like I really believe in this anti-to-do list. This is how I can like live my
4:04:09
independent person life with no teammates or employees, which as an extreme introvert is a little gift right
4:04:16
now. But like I don't do this enough which is and I and I try to do it a lot
4:04:22
which is every day when I have to do something miserable I think two things
4:04:27
before the thing I should actually think. The two things I think are I'm so busy. I'm so stressed out.
4:04:33
There's more to do than I can ever possibly get done. Oh my god I'm going to die. That's one. Second thing I go h
4:04:39
do I have to hire somebody? Those are the two first thoughts I always have. Those are not the right first thoughts
4:04:45
to have in this moment. The right first thought to have is like how can this never be a problem for me again and only
4:04:52
in this moment have I felt like that is actually a viable option for getting
4:04:57
stuff done inside a real company. And so I think notion is a good good solution
4:05:04
for this. But often people like forget all the things they can do. So I'm just going to give you uh we in sales call this an intimidation slide. It's like a
4:05:10
bunch of words that looks really impressive that no one will actually look at, but there's like a whole chunk
4:05:16
of stuff that you could just never ever ever have to do again or by yourself.
4:05:22
And I promise a lot of that is not the stuff that like brings you tons of intrinsic joy in your day-to-day. And I
4:05:28
really do think setting aside time and like being very explicit about burning
4:05:33
down this list, it's like tech tech debt for a burnt out human is really really
4:05:38
important. And people do not dedicate enough time to this. And then the last thing I would say from
4:05:45
an operating model perspective, so you have like workflows, you have the more tactical like let's get rid of the
4:05:50
mundane stuff. And then I do think like all and I'm going to go into culture in a minute, but like all operational
4:05:56
operating model changes, they get deployed through rituals. like they get
4:06:01
deployed through the touch points by which your team communicates, the touch points by which your team meets, the
4:06:07
frequency, the cadence, the place, all that stuff. And I think those rituals are are more important than ever right
4:06:14
now to get really explicit about and so you know this like 247 intern flow, how
4:06:23
do we make that visible? Does the intern work on Mondays? Like all these things sort of like create organizational
4:06:29
muscle around how stuff gets done. And I think rituals and making this systematic
4:06:35
inside your organization is a really effective way to reinforce this new operating model which honestly I think
4:06:42
can change how people do work. And I use that phrase a lot which is like muscle memory. We all have so much organization
4:06:49
muscle memory. Like when I need something prioritized I ask the PM. when I need a blog post written, I ping
4:06:56
marketing and like it is very hard to break that muscle. It's just like the thing that you do. And so the more you
4:07:03
can create alternative rituals, um the more you can actually change the way
4:07:09
your team operates and work gets done and it can often free up some like really creative thinking, really fun
4:07:15
work and a lot of joy. Okay, so the the second most important
4:07:21
thing is culture. This is the hardest part of all of this. I think culture is
4:07:27
honestly at the root of why people don't feel confident in this moment and I
4:07:33
think there's a lot of very specific things that you can do to improve culture, transform culture, change
4:07:38
things for your team. So the first thing I say is you have to have permission to
4:07:44
experiment. This is a moment of like high rate of experimentation and
4:07:50
honestly high rate of failure. So I'm just going to ask a question. We'll just
4:07:55
take it. It's Thursday, so it's a week. Who here, again, I was told to wake you up post lunch, so we're going to do like
4:08:01
the hand raising thing. Who here has been disappointed by AI this week?
4:08:09
Okay, we've we've all Yeah. Who here has like said like, "No, that's me. I just
4:08:14
go no to AI this week. Every single one of us every single day if we're embracing
4:08:20
these tools, if we're using these technologies are like deeply disappointed with them on a regular
4:08:25
basis and um our shame is hidden in like the sidebar of cursor. It's like I would
4:08:31
never talk to an intern like this but dear god my my desperation and sadness going to be here. And so much of what I
4:08:39
do it's going to change your life. It's so great so fun. and you can build stuff every you really need to embrace an
4:08:46
experimental mindset and a failure mindset and a success mindset and you
4:08:52
know as much as we all claim we're like go fast break things we love to experiment blah blah blah people do not
4:08:57
like failure they do not like sharing failure they do not like being a naysayer they don't like saying this
4:09:04
doesn't really work but actually in this moment where there's real impactful transformation happening the more you
4:09:11
can make experimentation truly viable in your in your organization, the more you're actually
4:09:18
going to discover where the leverage is. And so I do think this is a place where our risk tolerance has to get a lot
4:09:24
higher. And I'm not saying like risk tolerance for the business perspective. I'm honestly saying like us as individuals, little human hearts, our
4:09:31
little team hearts, we need to get a lot more comfortable with sharing failures
4:09:36
and successes and being willing also to try. I think so many of us high
4:09:42
achievers, I mean, you're like at a conference spending your time getting better at your job. So many of us are
4:09:48
very comfortable around mastering something and then like showing off our mastery like that's a good that oh that
4:09:55
feels good at work like I'm going to get promoted. I'm such a pro. And I think
4:10:00
having an organization that impre um embraces a novice mindset where it's
4:10:05
like actually we're all new here. We don't all have to be masters. We'll give
4:10:10
a lot more of the team permission to learn and start and it can make a really big difference.
4:10:17
And then there is like the actual practical uh uh permission to experiment
4:10:23
and uh but somebody told me they did this and I was like and I recorded a podcast this morning and they brought
4:10:30
basically this slide up and I was like this is the first thing I tell people to do. so boring and is like table stakes
4:10:37
which is permission to experiment also has to come from GNA our friends over in GNA um which means that like you have to
4:10:44
have money you have to have structures and you have to have permission to do this and so many teams are not confident
4:10:52
about their adoption of AI because literally they do not know am I allowed to how do I get this will
4:10:59
anybody pay for it so the foundations of an experimental mindset are actually found in our friends finance legal and
4:11:06
security which is one you have to define the golden path for AI tool adoption
4:11:11
like you have to be like if you want Bob's AI codegen CLI agent that just
4:11:18
came out yesterday that everybody's saying is hot like who do you ask how do
4:11:23
you try who's going to pay for it and then you like got to get it paid for in a week so that there's acceleration and
4:11:29
so I just think legal finance um policy being really aligned on like what is our
4:11:34
structural experimentation framework very helpful uh two you have to build in public by
4:11:42
default and I think this goes back to like we're all so comfortable in like showing off at work um but I think
4:11:47
getting into this mindset of like sharing failures and successes in public
4:11:53
constantly somebody called this like extreme many to many sharing which is
4:11:59
very very public very large channel and a high amount of volume of sharing right
4:12:05
now is going to give you organizational confidence and context about what's working and what's not working for your
4:12:11
team. And then uh failures or successes. So again, no I told you so. No like
4:12:18
engineers being like I told you that vibe coding code was crap. Like no designers being like see it's all AI
4:12:24
slop. Like we're not we're not doing that right now because it's not helpful. it does not help us get to where we want
4:12:31
to go. And what I tell people who love to live in this, love to naysay is I
4:12:38
often take them and I say, "Look, in two years when you're ready for your
4:12:44
next job and you're interviewing, do you want to be the person that figured this out or the person that said I told you
4:12:49
so?" and they're always like, "Oh, I got to be the person that like has figured figured this out because everybody
4:12:54
recognizes the inevitable, but there's a lot of stress in the immediate." And so I do think from a cultural perspective,
4:13:00
these three things go a very very far way. Uh the the other thing I think and again
4:13:07
um maybe this is like a just a reflection I'm a lonely bootstrap solo founder so you can take this you can just like throw this out if it's not
4:13:13
applicable to you but I do think operating as a team and recognizing
4:13:18
these techn the best analogy for these technologies now and I know it freaks everybody out is as an extension of your
4:13:25
team and I uh there's a lot of reasons that make me stress to share my Slack on like a Zoom a screen share the number
4:13:32
one is like all my friends are agents um and I have no one else to talk to. But I think if you think of these tools
4:13:40
as an extension of your team, not as tools used by your team, but an extension of your team, you can include
4:13:46
them in how you design your culture, how you design your rituals, how you design your operating model, how you design
4:13:53
your rewards, how you design all these things because they operate much more like interactive teammates than they do
4:14:01
like static tools. and and so I just think considering this all at the
4:14:06
aggregate level is a very helpful cultural mindset. Uh but again I'm just a lonely lady with six six agent
4:14:13
friends. Okay. And then the last thing I think you need to do from a cultural perspective and this is really important
4:14:19
in terms of driving internal confidence is you have to elevate your talent. This
4:14:25
is such a moment for talent elevation. This is like career making moments right
4:14:31
now. Um, and so so rare is there like an external force that can like really
4:14:37
change the trajectory of your career and this is one of them. And I think so many
4:14:42
folks get really focused in like what they can build and what they can do. They don't spend enough time inside
4:14:49
their teams pitching like how this is actually really really really fantastic for people and how it's really really
4:14:56
elevating their most talented highest impact people. And so I think a lot of
4:15:02
you know there's confidence from a am I going to ship good stuff. There's also confidence of like are we building the
4:15:08
team we want to be on? Do I have a future here? What's my job going to look like? Is this worth it? Like is this
4:15:15
going to be fun? And I do think this is a moment culturally where you really need to elevate talent for two reasons.
4:15:22
You need to elevate the voices of the people that are doing all this experiment experimentation work and are
4:15:27
leaning in and learning. You also need to elevate your talent. Literally elevate them out of the drudgery of
4:15:34
doing the work that they don't really enjoy. I tell designers all the time. I was like, you were not put on this
4:15:41
earth. Maybe, okay, honestly, maybe some of the notion designers were, but like you were not put on this earth to like
4:15:49
lay out beautiful forms. Like I love you, but like what if we took your craft up a little bit? What would you think
4:15:55
about if that wasn't your problem today? What? Like did you go to school cuz like
4:16:00
oh man I really wanted to like round that corner in Figma? No, you did not. Like you have such a higher opportunity
4:16:07
to have impact on your craft and if you can embrace these tools you can literally like elevate what your work um
4:16:14
is and I think that's very very valuable. So elevate your talent. I also think this is like career making magic
4:16:21
right now in in the next couple years and so if you lean in you can have a lot of opportunity.
4:16:27
Okay, so we're all gonna do it. We feel great about it. This pink has like hit all your dopamine centers. But wait,
4:16:34
like there there are things there are things to think about as you approach this. And like what are some of the
4:16:39
guard rails or warnings that I would give people? Um because there are traps.
4:16:45
These are the traps I say not to fall into. Not now, not me. Like good for you, not
4:16:51
good for me. This is actually the lecture I give CEOs, man. They love writing those emails about how AI is
4:16:56
going to transform everybody's job but theirs. Um, and so like I just think denial, waiting is the riskiest move. So
4:17:03
you need a forcing function. And that forcing function can come from operating model. It can come from culture. It can
4:17:09
come from a scary email. But like you need a forcing function. And a lot of people are like it's not yet for me. And
4:17:17
you can it's not yet for me. And then wait and then 18 months later be like whoops I missed. I I'm not on the edge.
4:17:24
And so I think denial is a big one. Second thing is secrets. There are still
4:17:30
so many se everybody's keeping their secrets. Um and again it's because
4:17:35
people like to look really impressive. And so I think one of the traps that I see right now is there are actually
4:17:42
really great pockets of AI adoption. Great people are doing amazing things and they tell no one and they don't know
4:17:48
how to teach each other. Like they actually don't know how to teach. I don't think it's uh ill-intentioned. I
4:17:54
think they actually just don't have the scaffold by which their knowledge gets
4:18:00
seated into the organization because they have no interest in doing that work on behalf of the company and so you have
4:18:06
to do it for them and so I think great ideas go undiscovered. You have to be really really intentional with your
4:18:13
frameworks for sharing. And then uh vibes only. I would say like
4:18:19
again a lot of orcs are like if I give them tools and I give them permission we
4:18:25
can all just through some tokens and like life will happen well. And I actually think systems are what scale
4:18:32
things through organization. So like systems are really better than vibes.
4:18:38
It's one thing you know like I hear this a lot from engineering leaders. I was an engineer like we should use AI to
4:18:44
accelerate our product roadmap like not actually like here's how you build a
4:18:50
cursor rules to burn down that tech debt that we have in our noisy front-end tests like there's just a gap between
4:18:57
like you can vibe your way there and you can systematize your way there and I think a lot of that is actually hard
4:19:03
skills gaps and so I've been lecturing people a lot on like this is the era of hard skills you actually have to have
4:19:09
like technical and tactical skills to be able to break down these tools into impact. And so I think it's like really
4:19:16
imperative that you don't stay up to date on the sound bites of which I hope this this this uh presentation becomes
4:19:23
one like you actually need to know the best practices. You need to know how to use these tools to their advanced degree
4:19:28
so you can make systems for them not just uh vibe vibe through the day. Okay,
4:19:34
so that's my pitch. Three things. Operating model, culture, avoid the traps. What does this mean for you and
4:19:40
your team? You're going to be competing with AI AI native teams more and more. Like it's
4:19:47
just the reality is you used to compete on like new product innovation ideas.
4:19:54
Now you're actually competing on the operating model of your team. It's like a like ground game is different now. And
4:20:00
I think you really need to think about what that means for how you show up to the market. It's really important. So in
4:20:06
response you need to create like you need to craft a strategy to answer that question because could be higher margin
4:20:13
could be higher velocity could be a lot of things but it's going to be different than what you're doing today.
4:20:19
You need to design your org operating model and culture with intention. Like this is not a moment where you're just
4:20:24
like this is going to happen to us. We'll see what happens and it'll be kind of the same but my SAS bills will be more expensive. Like you actually need
4:20:31
to design your organization. What role? what structures, what ratios, who goes
4:20:36
where, who has what, how do we get work done end to end with intention. And this
4:20:42
is a place where, you know, sometimes I don't we might not have a lot of like HR people folks in the room. You have such
4:20:49
experience in organization design like this is actually a moment where that's going to intersect with technology in a
4:20:54
really interesting way and create some cool cool jobs. And so every team should have an AI stack and strategy. And then
4:21:01
the last thing again is there's just no way way better to learn than getting hints on. I'm going to tell you like a
4:21:06
really dirty secret running a podcast about AI. I reach out to guests who have
4:21:12
a lot to say about AI and I go great. I would like you to come on and screen share three things you do with AI and
4:21:17
they're like what? Yeah. Like I kind of chat GPT every now
4:21:22
and then but I don't actually do anything. Like this is a moment where hard skills really really matter. I can
4:21:30
come up and lecture. I can tell you the world's going to change. Actually, what you need to know is like how do you prompt things? How do you do uh you
4:21:37
know, how do you find the right tools? How do you workflow these things together? And so, again, hard skill
4:21:42
development in this moment is really critical, especially for the execs out there. Okay, so I have a couple
4:21:48
challenges I'm going to leave you with. One, design the AI native version of your team or company. like actually
4:21:55
write down be like if what I believe to be true or what Claire believes to be true, it's crazy lazy with the agent
4:22:01
friends believes to be true is true, what is our team going to look like? How is that going to be designed? And
4:22:06
actually do that and then work backwards on the plan to get there. It doesn't have to be like today we change
4:22:12
everybody's title and no one talks to anybody but an appment mention in Slack. Like we work backward from plan to get
4:22:18
there. The second thing, this is a very fun one. You should actually do this every week. Write down your personal
4:22:25
anti-to-do list and then just like spend an hour a day burning burning that down.
4:22:31
Hold me accountable to it. I'm I still fall into the like who do I need to hire? I'm going to fail. I'm so stressed
4:22:37
out. When actually like this is the the answer. And the other good thing about this is it really is like just a very
4:22:45
practical, highreward, high cycle way to build hard skills. And so this is one of like the most practical things I could
4:22:51
recommend you do. And then my challenge three is like have fun. This is truly
4:22:59
truly truly truly the most fun moment I have had in multiple decades of being in
4:23:04
technology. I just think like I get to do more fun things with my time. I get to learn more things. I get to touch
4:23:10
more stuff. It is just a delight. And so I think one of the cultural gifts this
4:23:18
moment can give teams is like a moment to rembbrace the fun of creation and the
4:23:25
fun of learning new technologies and the fun of like actually being pretty wowed
4:23:31
by what's possible right now which is a feeling I haven't felt in a really long time. So I've never had more fun. I
4:23:39
think like there's this joyous moment. Um, and I think if you can make that like a scale cultural moment, you can
4:23:46
have a really high impact. So that's it. That's how you ship with confidence in AI. Please say hi to me. I really
4:23:52
appreciate it and enjoy the rest of your day.
4:24:02
Now take a 20 minute break in programming.
4:24:13
Fine. Yes. Okay, good. Am I going back here to getting Yeah. D
4:24:19
Mike this guy? Oh, hold on. Do you want this?
4:25:04
That was cool.
4:29:46
me.
4:29:51
Back.
4:31:04
Hey.
4:31:55
Hey,
4:32:03
hey, hey.
4:32:08
Heat. Heat.
4:32:35
Hey, hey, hey, hey, hey, hey, hey, hey, hey, hey, hey, hey.
4:32:55
Data
4:33:16
Hey,
4:33:33
hey.
4:39:31
Hey.
4:43:07
Hey, hey, hey.
4:43:28
Hey.
4:43:58
Please welcome Ivan Jiao and Claire Wait Keller.
4:44:16
Hello. Hello.
4:44:24
Um, okay. Uh, welcome to the afternoon sessions and here with me Claire.
4:44:32
Thank you for having me. Claire, if you don't know about Claire, she is the creative director of Unico.
4:44:40
You can't hear me. Can Can the back dial up my voice a little bit? Yeah. No. Oh,
4:44:46
yeah. Too soft. Sound up somewhere. Can we dial up the sound? Sound check.
4:44:52
Check. Check. Can you hear me in the back now? No. No. Needs to be higher.
4:44:58
Volume up, please. One, two, three. One, two, three. Anybody here? Can you hear in the back?
4:45:05
No. Can you hear me in the back? No. They'll know. Okay. Okay. Let's try again.
4:45:13
One, two, three. One, two, three. Back. Go back. They'll know. Yes. Or up.
4:45:19
You can. Sarah, you're the only go up a little bit. Maybe
4:45:25
up. Up. Keep going. Up. Keep going. One, two, three. One, two, three. One, two, three. Anybody here?
4:45:32
One, two, three. Keep going up. Okay. I wish I had my phone here. This is kind
4:45:37
of funny. I feel like I need a slight little one, two, three. One, two, three.
4:45:43
One, two, three. Everybody's Can you hear me now? Okay. Thumbs up. Right. You want to try? Make Make sure
4:45:49
everybody can hear me. Yeah. Okay. Great. Okay. Okay. Now we start. Let's start. So, uh, here with me is
4:45:56
Claire. Thank you. Claire is the creative director of Unico. So that means is all the thing you're
4:46:03
wearing, touching, buying from uni decided by Claire
4:46:09
my me. Uh raise your hand if you are uh heard of a uni code.
4:46:15
Everyone raise your hand if you're wearing uni code right now. Thank you. Yeah,
4:46:22
you know at notion we always talk about there are type of tools that doesn't
4:46:28
matter it's the money you cannot buy. Yeah. There are things like your iPhone the
4:46:34
richest person or a student can use the same phone right here. My favorite t-shirt is Yinko
4:46:40
Suprea t-shirt. Thank you. Doesn't matter. I love that one too. Yeah, it's any it's literally the best
4:46:45
t-shirt in the world made from in my opinion. Uh so today we can talk a
4:46:51
little bit about your um your life before Uniqo. Um Claire has a deep history of in
4:46:58
fashion houses before Uniq. We can talk about that. We can talk about craft. We
4:47:04
can talk about uh technology with fashion because Unico is very much tech
4:47:09
for as a company. Sounds good. Yes. All right. Let's go for it. Uh I'll start with some photos. So let's
4:47:17
start with this one. your first job. First job exactly back in the 90s dating
4:47:23
myself, but my age is online, so I'm not too embarrassed. So, um, yeah, back in the early 90s
4:47:31
coming to New York and working at Calvin Klein right at the pivot point when he
4:47:36
was really changing from being kind of an American sportsear designer into what
4:47:44
became like the height of minimalism. And you probably have heard the term
4:47:49
heroine chic and the waif. Um and so he kind of coined that era from the work
4:47:56
that we did at that time. So I joined in 93 and it was New York in a you know
4:48:03
very sort of raw gritty state. It was um not the kind of polished city you see now but uh exciting time to be there
4:48:11
that's for sure. Was that the hippiest brand back then? It was the hottest brand. Yeah. in America.
4:48:16
It just came out of nowhere. Yeah, it it I mean he'd been sort of building it through his jeans and
4:48:22
underwear and I think Oh, it's so interesting. Underwear first. Yeah. So, jeans was the big first. I
4:48:27
mean, he had some ready to wear obviously, but when he hit on the idea of designer branded jeans, there were a
4:48:35
few women that did it before him, I will say, Gloria Vanderville and there was another one, but he really marketed it
4:48:42
in such an clever way and made it super desirable and sexy um through Brook
4:48:48
Shield's ad. And then from that, he said, "Well, what else can I brand?" And
4:48:54
in a way he was one of the first to do branding and that's and was a really interesting experience for me being in a
4:49:01
company at that time when he was discovering that how powerful branding can be.
4:49:07
That's interesting. So underwear is sort of a commodity type of thing something you never see. Right.
4:49:13
And jeans is like at that time like Levis's Gap already popped. Exactly. Those were the ones but they're not fashion items.
4:49:19
Exactly. So his whole concept was how can I make this cool and super desirable
4:49:25
and he's his name was becoming more and more strong and so he chose controversy
4:49:31
as his route to marketing and so the big you know there were two big things he
4:49:36
did was the Brookshield ad which was banned and then um when I arrived at
4:49:42
Calvin within the first year he launched something called CK and CK um had a
4:49:48
perfume that came out CK1 and this fragrance he used complete street
4:49:54
casting from New York and it everybody looked like they were from downtown Soho
4:50:00
like artists, drug addicts, like very very sort of you know uh different to
4:50:05
what the luxury branding had been in that era. And so it got splashed all
4:50:11
over the news. And I remember coming in the office the next day and Calvin said,
4:50:16
"Oh my god, it's the best thing ever. we got banned everywhere. Um because he knew that controversy
4:50:24
explodes your brand. And so, you know, good news or bad news, it's it's all good.
4:50:30
Why didn't people figure it out before until the '9s? This is like all the
4:50:35
ingredients are here. It's true. I don't know. I think high fashion really at that stage was very
4:50:41
elitist and very pure. And so he felt there was an interesting route to reach
4:50:48
more people through less expensive product. And so the idea to bring
4:50:53
something like underwear, which you know, like I said, you normally never see, he said, well, how can I make this
4:51:00
incredibly sexy and desirable? So he hired Marky Mark. And then, you know, that, you know, that image from the '9s
4:51:08
of Marky Mark propelled everyone to start making branded underwear. I mean, it was like such a pivotal moment. And
4:51:15
so, I mean, I remember Marky Mark and Kate in the office and them hanging out.
4:51:20
They were both like late teenagers. It's kind of an amazing time to be there. Wow. Okay. 90s New York. It's if I can
4:51:27
live another city in a different era. It's a cool London. ' 90s New York. Right.
4:51:32
Definitely. Um, fast forward a little bit. We're gonna be jump back and forth in different pictures today. Yeah. Um after
4:51:40
that you went to Ralph Lauren and started their purple label, their high-end.
4:51:45
Yeah. So totally a different um aspect of design. So going from women's wear to
4:51:50
men's wear. And what was so fascinating about that was really the fact that I was coming in with Ralph um personally
4:51:59
him involved to start this new area of the brand. So it was almost like having
4:52:05
a little startup within the giant corporation of Ralph Lauren. So there were only four of us on the team and we
4:52:11
were working directly with him onetoone and just spending hours. I mean I I
4:52:17
would easily have a 5hour meeting with Ralph Lauren like sitting there at a table just like looking at fabrics and
4:52:25
concepts and color and him just talking and talking and talking. It wasn't like the case with CK. No, I
4:52:33
think you know with Calvin he we would do reviews with him but it was he was
4:52:39
already kind of on his path because this was a personal startup for Ralph. This
4:52:44
was something he'd always wanted to do was create this very high-end luxury men's wear. It was a personal project.
4:52:51
So he wanted to dedicate a lot of time to do it. Yeah. And um I read somewhere you mentioned
4:52:57
that cuz later on you went to European houses. Yeah. compared to American and European ones, Americans a lot more driven by
4:53:04
consumerism, commercialism. Yeah. How would you balance that even back
4:53:10
then between creative freedom and what's what sells? Yeah. It's it it was an interesting
4:53:16
dynamic. I mean, the two brands that I work for, Calvin and Ralph, have very distinct point of views and um strong
4:53:23
aesthetics. So we'd, you know, be able to kind of maneuver a little bit in terms of being creative, but it had to
4:53:30
be within the aesthetic of of the house. So it was a really important part of the
4:53:36
job was really understanding that because actually that's the power of a brand. And if you can recognize
4:53:43
something without even seeing the branding, that means your brand is incredibly powerful. It transcends
4:53:50
language. Yeah. And so that's something that would totally be immersed in
4:53:56
constantly sort of referring back to does this look very Ralph is is this something you see someone standing there
4:54:02
you're like oh he's wearing Ralph. It's kind of like the Coca-Cola bottle you can just the shape without you know
4:54:08
it's Coke. You want to drink Coke. It's like instilling the DNA in every part of the uh the clothing. You
4:54:15
must be quite young back then, right? Before I was 21 when I joined. Yeah.
4:54:21
And then Ralph Lauren, you starting a label for Ralph like how where were you
4:54:26
before that? How did you learn like were you in school? Yeah. So I I had um two I mean I I had
4:54:32
two sort of pivotal moments. When I was in my like early teens, I realized I really wanted to get into fashion. So I
4:54:39
decided to immediately specialize at 16. And so I went um then and did five years
4:54:46
of a sort of bachelor's degree and a master's degree. And my master's degree at the Royal College of Art was
4:54:51
specialized in um technical knitwear. So that meant industrialized knitwear. So
4:54:57
it was a it was a very very niche role but it meant that actually I had like
4:55:03
this incredible skill that nobody else had. And so when I graduated,
4:55:09
honestly the the landscape of the job market at that time was quite bleak and I got a job before I even um submitted
4:55:16
my graduate collection. Wow. It's kind of amazing. And they all trust you to start your own sub label in the
4:55:21
Yeah. know back then it's kind of interesting because we were really young but that kind of sense of youth and
4:55:28
bringing all of the sort of I don't know street culture and our sort of touch
4:55:33
point on what was happening was something that especially Calvin really wanted in the business Ralph was much
4:55:40
more about the fact that I was British and I could bring a lot of that Angloile
4:55:48
kind of uh knowledge to him you know history and culture from the UK was a a
4:55:53
big part of his brand. Cool. Cuz I I'm always fascinated by like, okay, if you spend five, six year
4:56:00
in college, you learn all the craft. Yes. of fashion. And is that enough? Then from there, it's just all improvising
4:56:06
for you. Yeah. And and the craft is really about that idea of understanding physically
4:56:12
how to do it. So being trained how to cut something. Um in my case, it was knitting. And also then um and knitting
4:56:20
is not the knitting needles. It's an actual industrial processes with like all this uh let's say kind of
4:56:27
programming that you have to do on the on the machines. Um and then from there
4:56:32
really understanding how to sort of design and and the skills that you sort of learn in in college are really about
4:56:39
kind of expanding your um experimentation. And so pushing boundaries really trying to understand
4:56:46
how you can create something new and as a as original as possible but within
4:56:52
within a framework and then when you go into companies and I think that's why ultimately I left America was because I
4:57:00
felt that I wasn't pushing the boundaries enough. Mhm. Okay. So it was it was a great learning curve.
4:57:07
I mean there's no better place to learn than America in terms of getting your sort of um skills as an actual designer
4:57:16
to sell product. Like I said the commercialization which is a very critical part obviously of business but
4:57:22
it's also um branding branding and marketing is really well done in in the states. So I find that that's those were
4:57:30
the two skills I sort of walked away with. I guess American back then there has new vacant space. So new brand can
4:57:36
pop up where European has all the houses. Yes. Yes. Exactly. If we move quickly to the Europe now. So
4:57:42
if we jump uh to here's Gucci. Yes. In how do you pronounce it? How would
4:57:47
you pronounce it? Gucci. Okay. Great. Gucci. Um in Milan and Chloe in Paris, I
4:57:53
believe. Paris. Yeah. And uh here's Givvanchi in Paris.
4:57:58
Yeah. Which city is your least favorite among the three?
4:58:03
Um, for what reason? At the time, I think Milan was quite tough, I have to say. Um, it's a very
4:58:09
small city and back then that was like 2000. So, um, there were really
4:58:16
five or six key big leading brands like uh, Prada, Gucci, Versace, Armani,
4:58:23
uh, I would is there another one maybe that was key back then? Um but anyway
4:58:28
they we really had um a kind of a very
4:58:33
small office there and so I was going between London and Milan constantly and
4:58:38
there wasn't a as much vibrant scene happening in Milan. It felt very much that uh Paris was where everything the
4:58:46
center was building but for me actually being in in Gucci at that time was quite
4:58:52
an exciting moment to join the company. I joined when Tom Ford was working there
4:58:58
and he had started to really build this huge momentum and when I joined in 2000
4:59:04
it was just like going on a huge trajectory and that's when a year later they started the Gucci group which is
4:59:11
today now called caring um but at the time was didn't even exist when I
4:59:17
joined. Nice. Let's quickly jump to where you are today. Yeah. which is uh
4:59:24
here you are unico uni that's I think that's your line right that's yes this was a collection that came out
4:59:30
in last year yeah it's so it must be quite different than the European houses and the American one
4:59:38
before like how would you what are the tradeoffs that you need to make in uni
4:59:45
code that you don't have to consider before it's it's interesting because there are things I trade off such as obviously the
4:59:52
high luxury and also there are certain things that you would never find at Uniqlo like you would you wouldn't find
4:59:58
a sequin top but you you won't find an embroidered dress at Unilo because it's just not their language um and not also
5:00:06
what they stand for so those things that came out of my previous roles in Paris
5:00:12
especially Khloe and where is working in high couture and working with like
5:00:17
incredible artisans in Paris beautiful atilier days. Um, that's something yeah
5:00:23
I have to trade off. But what I've gained is something also very interesting which is working in Japanese
5:00:31
culture and also the technology and the technology part is really a huge bit of
5:00:37
the the success of of Uniqlo and you you what you gain is everybody
5:00:42
can raise their hand here versus exactly versus not everyone wear Chloe here. No, that's the thing. And I I thought it was
5:00:48
very interesting throughout my sort of 10 years in Paris, I'd had so many
5:00:54
people who come up to me and say, "Oh my god, I love your clothes, but I just I can't afford them. Maybe one day I get
5:01:00
one piece." Um, and now everybody comes and say, "Oh my god, I bought five pieces from your collection, and look,
5:01:07
I'm wearing this." Um, and it's a really different dynamic, and it's it's super rewarding. Is it more rewarding than
5:01:14
creating something so delicate and beautiful but fewer people can? How what's your calculus between quantity
5:01:22
and qual not quality is wrong quantity versus I mean I've been lucky enough to have
5:01:27
both. So I would say um it's interesting because I would
5:01:33
definitely say the couture was like one of the highlight experiences of my career. It's a it's a a super beautiful
5:01:40
part of French fashion and it's an area where you really get to completely be in
5:01:48
almost like your own laboratory. It's it's about kind of shutting yourself in a room with um just really incredible
5:01:56
artisans of embroidery people who do beautiful like um special printing
5:02:02
techniques. Um we did amazing jewelry, bespoke shoes. I it's like a dream
5:02:08
world. It's kind of something you you cannot explain. It's it's it's really a rarified world. Um it's the top of the
5:02:16
top in in so many ways. And I created you know uh red carpet dresses, wedding
5:02:22
dresses, all these like extraordinary pivotal moments like iconic moments on
5:02:28
uh red carpets and things like that which uh will live in my memory forever.
5:02:33
They were extraordinary. But at the same time, there's something really dynamic about being able to dress the world, and
5:02:41
that's something that I equally find super inspiring. Going back to the word tradeoffs, like
5:02:48
now you're creative director of one of probably the most popular clothing brand
5:02:54
of the world. Yeah. And you what do you need to balance? You need to balance the the fashion, what's
5:03:01
going on in the season, uh materials, technology, business, how
5:03:07
much. Yeah. Like what are the dimension of trade-off you have to consider at your There's a lot of tradeoffs.
5:03:12
Um there's a lot of trade-offs because of price obviously. So there are certain limitations on that. But also um there
5:03:20
are trade-offs in terms of sometimes I I need to prioritize function over pure
5:03:27
aesthetic. And that's something that I have to give us one example. So an example for instance I had this
5:03:34
like beautiful coat I was working on last season and I had added extra sort
5:03:40
of um pocket and zip detailing because I just thought it looked great. And when
5:03:45
we were going through the motions of how we were going to produce this item and always it's a discussion with the
5:03:51
founder Mr. and I um and I sit together and um he's like, "Is that really
5:03:57
necessary?" And I said, "It's not necessary, but it
5:04:03
kind of looks great." He's like, "Okay, just think about it." And always when he says that, I know what he I know what
5:04:10
he's getting to. He's If you're trying to reach a global customer, you don't
5:04:15
want anything that's going to put somebody off buying it. So, is that pocket or is that detail going to be
5:04:22
irritating or oh, I love this, but that's kind of annoying that's there. And so, I have to make those decisions.
5:04:30
And in the end, I did keep it. I just tweaked it a little bit. Um, but it was
5:04:35
something where it you they have this lovely phrase in Japan where they say,
5:04:40
"We need to think deeply about this." And it's a it's a phrase we use
5:04:45
constantly in meetings. and they they genuinely do spend several weeks
5:04:52
thinking deeply about it and then we come back around and discuss it and say okay maybe we will make that change or
5:04:59
maybe we won't. That's fascinating. So different does unique code or do you have to change the
5:05:04
the the the fit to different regions differently because culture might accept things different. Definitely. um examples. It's complete.
5:05:12
It's a completely new way of working for me that because when you're a luxury brand, you really are um just designing
5:05:18
for a runway. That's your whole goal is to get a message across on a runway and that's your like marketing for the
5:05:25
season. So that's what's going to sell your concept, your idea, what you think is the direction of fashion for the next
5:05:31
six months. Whereas at Uniqlo, I'm really thinking about, okay, what are people going to wear every day? what do
5:05:37
we need to complement what we already have in the collection? Um, and so we're we're really thinking about how we
5:05:45
design also for different regions. Of course, Southeast Asia is a tropical climate. Europe is cold most of the
5:05:52
time. So, we slightly adjust the way we kind of design certain parts of the
5:05:57
range to accommodate for those kind of especially climate differences. Climate is a huge part of the work we do now.
5:06:04
Wow. Yeah. No wonder some t-shirt feels better here versus in Asia. Right.
5:06:10
There are there's different fits as well. Um we do adjust everything in terms of regional differences. So Asia
5:06:17
has a slightly smaller fit and a shorter fit. But it's um it's interesting because I actually had to do a uh data
5:06:25
analysis of uh the whole global footprint of people's body shapes. And
5:06:32
it's really interesting to see kind of globally how it shifts around the world. And of course like Scandinavia is the
5:06:38
tallest. Um but then interestingly because of um you know over the last I
5:06:46
don't know 50 years America has got a lot of Asian immigration. So actually
5:06:52
it's a similar fit to Asia. Um and I thought that was kind of fascinating because I did not expect it. And so it's
5:07:00
um something where when you see and we we work with um big institutes to get all this sort of global information but
5:07:07
it's something I never would have done in my previous roles. Fascinating. I thought American also getting bigger too.
5:07:13
Yeah, it did go kind of that and that way. H interesting. And Unico is famous for
5:07:20
their collaborations. Yes. Raise your hand if you are aware of the Unico collabs.
5:07:26
So most people know what they are. Yeah. And before you working inside other houses like Ralph Laurens, CK,
5:07:33
Calvin Klein, then you work for other people. Yeah. Now the collaborator probably need to work with you or for you.
5:07:39
Does it diff is it different craft, different trait to manage those other designers? Well, I mean Jonathan Addison and and
5:07:46
Christoph Laare who does the Uniqlo U line, they actually design independently. So I'm not overseeing
5:07:52
them as such but their their let's say their collaborations sit within um the
5:07:58
whole universe of Uniqlo. So when we do a review of what will go into the stores
5:08:05
then there are um considerations that you know there will be not necessarily
5:08:10
from me directly but from the merchandising teams that will go back and say okay we need to reduce the
5:08:17
collection a little bit here or there based on on how the balance of across
5:08:22
all the products that we have is and that includes my range as well. We we sort of assess so that there is no
5:08:27
duplication. And I think one of the things that they're very careful at doing is to really make sure that every
5:08:34
product has a purpose in the store. Like we're not doing a blazer in that collection and a blazer in that
5:08:40
collection and a blazer in that collection. If we're going to do a blazer, that's the one we're going to do. Who's deciding this?
5:08:46
Mostly Mr. Yai. It's I think I don't want to get involved. I mean, they're my peers, so
5:08:52
I, you know, we we're all sort of on on the same level in that sense. But, you know, the merchants along with Misty and
5:08:59
I will make the final call on the the actual I mean, because it's it's an
5:09:04
astonishingly huge business. So, uh, if you make a mistake, it's a big mistake. Oh, makes sense. And from spending time
5:09:12
with you this trip, I get to know that you're really f you're really interested in technology.
5:09:17
Yes. Right. Um, San Francisco AI. What What outside maybe venture outside
5:09:23
fashion a little bit? What interests you about it? You you seem to like it can span from couture to know.
5:09:29
Yeah. Underwear, not underwear, but Calvin Klein in the 90s to uh unique the present day and now other things like
5:09:36
what where where is your energy taking you? I've, you know, I've always loved the idea of moving a lot in my career.
5:09:44
And I think I've chosen really different brands every time I've moved. Quite unexpected choices a lot of the time.
5:09:51
And I think that's part of the reason I do that is because I like being um challenged and being in an uncomfortable
5:09:58
zone. I think that really forces creativity because you end up really
5:10:05
fundamentally changing your view of something. And I think as a designer and a creative, you have to constantly
5:10:11
refresh yourself, reinvent yourself because otherwise you get stale and then it just becomes not relevant. And one of
5:10:18
the things you learn about in fashion is that the industry changes so rapidly
5:10:23
that you have to be constantly on the edge of it, constantly in front of it. I
5:10:28
mean, I'm currently designing spring summer 27. So that's like 18 months, two
5:10:35
years ahead of where we are. and I have to kind of be able to predict in like a zeitgeist way of where that's going and
5:10:42
know how to navigate and um do it. So I I like the idea of challenge and I like
5:10:48
the idea of being in areas of creativity that are are new and exciting.
5:10:54
Let's uh dive into that a little bit. So you need to design 18 months from now. Yeah. How do you solve this problem?
5:11:02
It's a puzzle. Yeah. What's what are the pieces and how are you thinking about it? It's um it's a lot of like gut instinct
5:11:10
which I know is a hard thing to describe but it it's true. There is a especially as a creative you have real strong
5:11:16
instincts for things and some of that is honed over many years of just knowing. Um, also it's about understanding
5:11:24
cultural shifts, understanding what's happening on the street, understanding what's happening in younger generations,
5:11:32
uh, understanding influences in the arts, in theater, um, you know, you you
5:11:38
see it happening all around you. And it's my job to kind of pick it up in
5:11:44
different places and start to find similarities, start to find, let's say,
5:11:51
things that I I'm starting to link together. And so I I remember talking to
5:11:56
Missy about this, the founder of Unilo, and he's like, "How do you know? How do you know?" Um because I had a a really
5:12:03
huge success last September with this like sweat hoodie and sweatpant that just it's actually the one this this
5:12:10
sweat hoodie here. Okay. Blew up everywhere. It sold out three times within like 2 or 3 months and um
5:12:16
we've just like constantly been restocking it. And he's like, "How how did you know? How did you know?" And I'm
5:12:22
just like, "I just knew." Um but I also knew what shape would be right. I knew
5:12:28
the fabric had to have something new and modern to it. Um I I kind of know from
5:12:35
my luxury background sort of what design where design is sort of moving to. So I
5:12:40
wanted to bring an element of that luxury instinct that I had of where things were
5:12:46
going into the pro this sort of quite everyday product. And normally you would find that in like a classic
5:12:52
sweatshirting but I did this in a very specialized doubleace fabric. Um, so it
5:12:58
created much more structure and the silhouette was more relaxed. It was a
5:13:04
slight different proportion and so it ju it just exploded everywhere. And those
5:13:09
are the kind of things I have to pick up on is like what are those essential pieces that have like exactly the right
5:13:17
feeling for the time? Do you believe this like because compare what you just had?
5:13:22
Yeah. versus the mid late 90s Calvin Klein because there's similarity of
5:13:28
similarities. Are you do you are you the believer of fashion runs in 25 years cycles?
5:13:34
It does. And I know why is that? It's some kind of universe calling of this. I I don't know. There's just some I
5:13:42
think there's a there's a lot of to do with generations. Um, and I think
5:13:48
certainly for me right now, I have two daughters, 22, and they are just rifling
5:13:54
through my ' 90s wardrobe, but there's no tomorrow. Um, and I still have quite a few pieces from that time. And I think
5:14:01
there's something about that sort of generational gap. And I remember doing it with my mothers.
5:14:07
And I think it's just when you like are in your late teens, early 20s, there's
5:14:12
something about wanting to dive backwards and like pick through like
5:14:17
something from the past that you can own for the future. I don't know. It just it always happens that way. I did it. Now
5:14:24
my daughters are doing it. I think it's just going to continually happen. But there's some there's a little bit of nostalgia there, but then there's always
5:14:30
a an element of wanting to twist it in a new way. And it's so amazing some designer who can ride a wave multiple
5:14:37
times. Multiple times like who are the great ones for you? Ah I mean I think well let's see. I mean
5:14:44
he just died actually but Armani um I think is an interesting one because he obviously you know he I mean he really
5:14:52
is the one of the unicorns of fashion. He owned you know right up until the day he died 99% of his company. It's kind of
5:15:00
extraordinary. um and he built an empire um from a suit and it and it just is
5:15:07
it's a an amazing example of someone who sort of really had a vision and just stayed completely
5:15:14
true to his vision the whole way. Ralph is another one. He sort of goes in and out of fashion. He's having another
5:15:20
moment right now, but there was a moment maybe 10 years ago where it was a little bit more quiet and um the preppy vibe
5:15:28
just was less relevant. But he didn't deviate away from it. He just kept going
5:15:34
and doubled down and just, you know, kept the image building even stronger.
5:15:39
And so now that it's sort of come back around again, he's like right at the epicenter of it.
5:15:44
Interesting. Maybe another way to ride a wave, just stay where you are and wait for 20 years, then the wave come back to
5:15:50
you. Yeah. I mean, you have to have some strong guts to do it for sure. But it's
5:15:55
true. Nice. Uh, thank you. This is great. We're learning a lot about the internal
5:16:01
working of building a fashion house and a creative brand like Unico. So, thank you for having time with us.
5:16:07
Thank you so much. Yeah. Thank you. Thanks everybody.
5:16:27
Heat.
5:16:45
Hey, Heat.
5:17:28
What?
5:24:00
Please welcome Jean Grosser.
5:24:08
All right. Hello everyone. Um, as you've just heard, my name is Jean Grosser. I
5:24:15
am the COO of Verscell and I'm excited to chat with you all today. Um, we have
5:24:21
had a lot of incredible presentations today and if there's one overarching
5:24:26
takeaway, it's that AI isn't a feature or a product. It's an operating model
5:24:32
that rearranges every part of how we run companies, how we make things, how we
5:24:38
define roles, ultimately how we create value. The companies winning in AI
5:24:43
aren't just adopting new tools, they're reimagining ways of working.
5:24:50
Both Fursel and Notion are in the business of enabling our customers to adopt these new ways of working. In
5:24:57
notion, agent turns messy inputs into structured work. With Verscell, V0ero
5:25:05
turns intent into UI in seconds. But we're also AI first businesses
5:25:12
ourselves. And you won't be surprised to learn that notion has played a big part
5:25:17
in that shift for Versell. I've spent my year working at
5:25:23
developer-led companies. First at Google, then a chief business officer at
5:25:28
Stripe and now at Purcell. I'm passionate about bringing novel
5:25:33
approaches to bear on go to market. At Stripe, that's centered around adopting
5:25:38
consumption-based business models and at Purcell it's reinventing go to market in
5:25:44
an AI first era. Over the course of my uh career, I've
5:25:50
found culturally that companies tend to skew towards one or two ends of a
5:25:55
spectrum when it comes to building. On the one hand, you have written cultures. Stripe was a classic example of this. I
5:26:02
think everybody knows Amazon AWS uh also famous for hardcore doc written culture.
5:26:08
And on the other end of the spectrum is a visual presentation. Uh it's this one is strongly embedded at Verscell and
5:26:15
other creative companies like Apple. I like to call this docs versus demos.
5:26:22
Uh and there are huge strengths and admittedly drawbacks to both approaches
5:26:28
and they don't often play that well with each other. As you can imagine, transitioning between these two cultures
5:26:34
can be a challenge. So, what if written cultures and visual
5:26:40
cultures could finally converge? As you saw today, they can when the doc
5:26:48
literally produces the demo. Most companies uh know that they need to
5:26:56
rethink their approach to software, the software development life cycle. At Verscell, we've already shifted to a
5:27:03
world where the lines between designers, PMS, engineers are increasingly blurring
5:27:10
and velocity is increasing. Now in its 10th year, Burcell is having
5:27:17
its best year yet. We've introduced the AI cloud, a unified platform for
5:27:22
building, deploying, and running intelligent applications and agents.
5:27:27
And we've seen 80% growth year-on-year, crossed over 200 million in ARR back in
5:27:33
the spring. And despite B 0 being in its infancy, over half of its customers are
5:27:39
large enterprises. We iterate and ship fast. And our shipping velocity is
5:27:45
driven by 16 different product areas. As
5:27:50
you can imagine, a challenge emerged for us in recent years. How do we maintain
5:27:55
our signature speed while ensuring quality and alignment across this broad
5:28:01
group of product stakeholders? Enter yet again notion. Notion has
5:28:07
enabled Verscell to create cohesive standardized systems that work for everyone. Product area plans help teams
5:28:15
stay aligned through a central roadmap that connects highle strategy to actual
5:28:20
work. Uh, and notion AI distills lengthy docs
5:28:26
into executive summaries and transforms technical documentation into customer
5:28:32
ready messaging drafts that my go-to market team can leverage.
5:28:37
Like notion, we're huge fans of quantifying impact. So, here you go. 35%
5:28:44
faster shipping, 89% confidence in quality, and nine hours saved per person
5:28:50
weekly by running our operating system in Notion. Notion AI uh AI through the power of
5:28:59
notion is giving our product teams a day back every week.
5:29:05
I imagine everyone in this room has some sort of AI mandate, but you may not yet
5:29:10
have a firm vision for how to deliver it. Today, I'd like to talk to you about how to successfully grow an AIdriven
5:29:18
business, drawing from Versel's experiences and those that we're learning and talking to a bunch of our
5:29:23
enterprise customers. We'll do this through three lenses:
5:29:28
people, practices, and product. more specifically, how to create AI powered
5:29:36
roles, contextdriven workflows, and smarter products that drive competitive
5:29:41
advantage through your company. So, let's start with people. It's time
5:29:46
to start hiring for roles that didn't exist a year ago. The first one is go to
5:29:52
market engineering. This was the first role I added at Verscell when I was less than six weeks into the company. I'm now
5:29:59
about six months into Verscell. Um, go to market engineering means that we
5:30:04
build our go to market like a product. We determine what we want the customer
5:30:10
exper the customer to feel like to be sold to by versel and then we
5:30:16
orchestrate those actions with AI at the foundation. So how do we do that? At a
5:30:22
high level we focus on three principles. We explore, we validate and incubate new
5:30:30
technical ideas from engineering, product and design into scalable go to market playbooks. We build,
5:30:38
we ship code that drives revenue with AI and data. And we redefine GTM. We dog
5:30:45
food the Verscell products and show customers how to redefine their businesses with agents, vzero, the AI
5:30:52
SDK and AI gateway. So to bring this to life in June, just
5:30:58
three months ago, we had 10 SDRs, sales development reps working on inbound
5:31:04
contact sales leads. Um, and you know, everyone on the team had directionally similar plays, but we did have one SDR
5:31:11
that was meaningfully outperforming the rest. Um, and so we paired that SDR with
5:31:18
a GTM engineer. that GTM engineer watched their workflow, looked at all
5:31:23
the data that they were working up, scrutinized the content, and then we were able to turn that in to an
5:31:29
agent-based play. So, it took us six weeks of a single GTM
5:31:35
engineer to build and refine our inbound agent. As of August 1st, we not only have just
5:31:43
one SDR, but also we are more effective
5:31:49
at uh qualification. We've gone from eight touches on average to get to an opportunity down to four. And our
5:31:56
opportunity conversion rate has stayed steady all along the way. And the other thing we did is we've gone back and back
5:32:02
tested everything and shown that when we disqualify an inbound lead, uh, we're
5:32:07
right 99.5% of the time.
5:32:13
We're moving with incredible velocity as we shift the go to market org to be AI first. And to make this happen, we've
5:32:19
built pods. So it works really well is you take your best functional role, you
5:32:25
pair them with a single go to market engineer and someone from the data science org and then they can codify the
5:32:33
best human practice, augment it with data, build an agent and iterate with a
5:32:40
human in the loop. The next area is practices. As I said
5:32:46
earlier, I think the company's best position to succeed with AI will be equal parts demo and docs.
5:32:54
As a new leader at Burcell, the single most striking practice at the company is demo days. I was truly blown away when I
5:33:01
started. So, every single Friday at 8 a.m. and this is an engineering, you
5:33:06
know, centric company. We do an hour-long demo day. These last five minutes, so we do 12 of them every
5:33:13
single Friday. The other thing that's striking is it's not just designers, engineers, PMs demoing. It's every
5:33:20
function at the company. Go to market, legal, HR, finance, all up there going
5:33:26
through what they're doing with Vzero with their prototypes there. This is also a great example of
5:33:33
where Notion brings two cultures together. Every demo has a notion page.
5:33:40
The agent summarizes outcomes and next steps and V0 embeds and preserves the
5:33:45
artifact. This is what we call contextual documentation. So notion not
5:33:51
only serves as our company memory, but it's injecting context.
5:33:57
And really, there's never actually been a better time to document every single
5:34:03
thing you do. All of that all of that documentation is what you ultimately feed into an agent. so that it learns
5:34:10
how to be best at its job. Who here has heard of context engineering? All right, lots of hands.
5:34:17
Good. Uh, hot topic in the AI engineering community, but it applies just as much to go to market. Let me
5:34:24
give you a concrete example of context engineering in GTM. So, let's say again, I'll use the STR uh
5:34:31
use case. We've sort of been starting here to try to enable it and or automate things end to end. Uh, so you got a lot
5:34:38
of STRs around the world. Most of these folks are college grads. Um, and this is their first job, but you want them to
5:34:44
reach out to an executive. In our case, that's often a CTO. And a lot of these folks might not have engineering
5:34:50
backgrounds themselves. So, suffice to say, my experience has been that when
5:34:55
they go and write an email that's meant to come from me to a CTO, it does not
5:35:00
exactly look like how I would have written it on their first try. Uh, so here's what we do. STRs go out and they
5:35:07
do research. Uh they learn about the company. They have a hypothesis about how our product would be a fit and they
5:35:15
draft emails in notion. Uh we then feed those into a Slackbot which I then go
5:35:21
edit. Um and then I I do this to match my voice um as well as have them be at
5:35:27
the right altitude. I send them then out actually to that CTO we're trying to get
5:35:33
in touch with and I bcc an email alias. Um and now we're able to take the
5:35:38
research from notion and the email content to make a gene agent. So after
5:35:44
30 days of doing this, we had this huge corpus of all the different emails that were gene approved that I felt
5:35:50
comfortable sending to an e-commerce company, a SAS company, um almost always in our case a CTO or CPO. Um, and so
5:35:58
that's helped our SDRs refine their content without me and essentially has now taught 40 SDRs uh how to talk just
5:36:07
like Gene every day. So we use it effectively as a coaching tool. They'll go and write an email and then the GAN
5:36:12
agent is like n didn't get this right like you know no executive reads a thousand words let's get this down to
5:36:18
20. Uh that type of deal. Um, but by the by Q4, we think there will now be a
5:36:25
portion of those emails that we have enough confidence to actually just go let the agent send.
5:36:32
Context now trumps data alone.
5:36:37
And while today we're using that system, as I mentioned, to train the SDRs to write like an executive, we don't think
5:36:43
we'll be doing all of that work by the end of the year. This is the power of context engineering. systematically
5:36:50
capturing not just what you do, but why you do it, how you think about it, and
5:36:55
what excellent really looks like. The companies that win will be the ones that are most intentional about capturing and
5:37:03
structuring your institutional knowledge. Finally, let's talk about products.
5:37:10
Uh Mark Andre once declared that software was eating the world. And I'm
5:37:16
here to say that I think that software with a fixed schema, so one that forces
5:37:22
you to adopt its workflow, is dead. Software behemoths have often touted, I
5:37:28
think everybody's heard this, for every $1 spent on our platform, $7 are spent in the ecosystem at large. And I would
5:37:36
purport that that was a bug, not a feature. Composable software and autonomous
5:37:42
agents will give us the per company and per user workflows that we've always
5:37:48
sought. This is where v 0ero and notion agent represent something bigger than
5:37:54
individual features. Notion or vzero has been labeled for better worse you know vibe coding. It
5:38:01
enables you to then go create that um peruser software and similarly notion is
5:38:08
in that by working for getting to more structured output.
5:38:14
I'm sure many people in this room are Salesforce users. How many of you have
5:38:19
ever sat at your desk and thought to yourself, "Man, this is exactly how I
5:38:24
wanted to do this type of work." Yes, exactly. said, "No one ever."
5:38:31
Um, so the reality is we've been stuck with these fixed frontends and workflows
5:38:38
that force us to work the way Salesforce or Workday or uh Service Now um thought
5:38:44
was best to work. And absolutely nothing wrong with any of those companies, right? I've used Salesforce for 20 years
5:38:50
at this point. Um, but these powerful systems will continue to be important
5:38:55
systems of record providing guards and workflows that are business critical.
5:39:01
But what's changing is that the marginal cost of custom software is trending to
5:39:07
zero. And that's going to usher in an opportunity for composable systems of
5:39:13
action. So, Notion should have its own Salesforce front end just like Burcell
5:39:20
should have its own Salesforce front end which is different because we have fundamentally different businesses
5:39:26
trying to achieve different things in our sales cycle. Retail figured this out years ago with headless commerce. Macy's
5:39:34
front end doesn't look anything like all birds, but they can both run on the same
5:39:40
underlying system of record, whether that's Shopify, commerce cloud, or something else. The rest of us across
5:39:47
all the other verticals are now just catching up to what retail has known for years.
5:39:52
We've seen incredible results from personalization in Verscell's e-commerce customers, from eBay to Elk Shop. Um,
5:40:00
and now we expect that same level of customization to carry forward into all software experiences.
5:40:08
Every company we talk to is somewhere on the AI maturity curve. Most have built
5:40:13
some sort of assistant, a chatbot or an embedded uh AI feature.
5:40:19
The next step is agents that can automate business processes or trigger workflows either to make employees more
5:40:27
productive or deliver better customer experiences. From there, the leap is to fully
5:40:33
autonomous systems that can orchestrate entire workflows endtoend, proactively
5:40:39
surface opportunities and make decisions across systems with minimal oversight.
5:40:45
As a AI workloads become a larger part of application traffic, as we move up
5:40:50
the curve, companies need to consider whether it's effective to build AI infrastructure themselves or engage with
5:40:57
a company like Verscell that specializes in it. And beyond your tech stack, I challenge
5:41:04
you to think about how how you work. To build an AI first business, your company
5:41:10
has to deeply understand best practices of your ways of working. You can't just slap an interface on top of a bad
5:41:17
process. AI is like building a house during an earthquake. The ground is always moving,
5:41:25
so you can't just build AI as a feature. You need to make AI your operating
5:41:31
model. Thank you everyone for having me.