Pigment’s Eléonore Crespo: Building France’s Next Unicorn With AI Agents And Tom Cruise

17 Jul 2026 · 43 min · 21 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

Pigment’s AI agents for enterprise business planning—replacing spreadsheet-heavy, cross-department finance/supply-chain/sales planning with real-time scenario modeling and faster decisions. The episode also covers “unreasonable hospitality,” co-CEO/cofounder dynamics, building in France for talent/loyalty, and managing AI token costs with model/usage governance.

Guest backgrounds

Éléonore (Eleanor) Crespo is Pigment co-founder and co-CEO. She started her career at Google as an engineer and later worked in venture capital (Index Ventures mentioned). Her co-founder Romain is described as former co-founder/CTO of Criteo, a French unicorn that IPO’d on NASDAQ in 2013.

Key claims

Spreadsheets can’t scale for strategic decision-making; Pigment’s agents connect data across teams and run dozens of scenarios. Analysts like the automation because it removes reconciliation/cleaning so they can focus on strategy. Pigment helps control AI token spend by comparing costs across models and allocating budgets across departments.

Notable examples

Customers named include Uber, Coca-Cola, Siemens, Unilever, Anthropic, ServiceNow, Datadog (token-management app built on Pigment), and Meta (cookies example). Crespo says she closed a CRO candidate after arranging a meeting with Tom Cruise via investor Iconic.

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

The Limitations of Spreadsheets

0:00 to 0:38

Learn about the inefficiencies of spreadsheets in data management.

“You know, they cannot help different departments communicate together.”

Eleanor's Background and Pigment's Mission

2:01 to 2:49

Explore Eleanor's journey and the mission behind Pigment.

“So Eleanor, we're in London today, but you're usually based in the Paris area, right?”

Challenges of Large Companies

2:49 to 4:00

Discuss how large companies face complex challenges and need agile solutions.

“What did those big companies want from Pigment specifically that they couldn't build themselves?”

Ideal Customers and Entry Points

4:00 to 5:24

Identify Pigment's ideal customers and how they engage.

“to be able to be more competitive, to be able to be more agile in this very complex environment.”

The Evolution of AI in Business

5:24 to 6:06

Examine the changing landscape of AI tools for business decision-making.

“They start where they have the biggest pain point.”

Data Complexity and Real-Time Processing

6:06 to 8:19

Understand the challenges of data management in large organizations.

“And what I saw at Google and that I saw as well at Index Ventures across a lot of cloud 100 companies is that whenever a CEO had to make a decision, they didn't have any tool to help them.”

Choosing a Co-Founder

8:19 to 13:00

Hear Eleanor's insights on the importance of selecting the right co-founder.

“And they need to run dozens of scenarios on it.”

From Ideas to Execution

13:00 to 14:00

Learn about the process of refining ideas into a successful startup.

“This book of ideas, were they all within the idea that the broad area that became Pigment, or were they different types of startups, and then you guys zoomed in on Pigment over time?”

Building Infrastructure vs. Prototyping

14:00 to 15:00

Explore the balance between speed and careful development in building AI infrastructure.

“You know, let's do something in the matter of days.”

Unreasonable Hospitality in Business

15:00 to 15:49

Learn about the concept of unreasonable hospitality and its impact on customer relations.

“You know, it's like I think everybody today can take some YouTube tutorials and tune a car and you can find every piece to make your own car if you want to.”
Show all 21 chapters

Delivering Exceptional Customer Experiences

15:49 to 17:41

Understand how Pigment goes above and beyond in creating memorable customer interactions.

“You know, when we come to meetings, the way we show up, the way we deliver the experience when they buy pigment, etc., we make it extremely memorable.”

Co-CEO Dynamics and Team Structure

17:41 to 19:13

Discover how co-CEOs manage their roles and foster a collaborative leadership environment.

“Well, I think, you know, first it's like we serve the enterprise.”

Clear Roles and Responsibilities

19:13 to 22:42

Learn the importance of defining roles to prevent conflict in a co-leadership structure.

“So I think, first of all, this complement is great and it's been a strength for us.”

Decision-Making in Co-Leadership

22:42 to 23:41

Explore how co-CEOs make critical decisions while maintaining their individual responsibilities.

“to also understand when, you know, the person might get upset if you don't ask them for their own opinion, right?”

Benefits of Building in Europe

23:41 to 25:41

Understand the advantages of establishing a company in France, including talent retention and cultural diversity.

“I knew that would be one for you for sure.”

Navigating Global Business Dynamics

25:41 to 28:05

Learn how Pigment leverages support from the French government while operating globally.

“Like, clearly, a lot of the innovation is happening there.”

Navigating Enterprise Sales in Europe

28:05 to 29:57

Learn about the challenges of selling to conservative European enterprises.

“And so for me, it's really important to juggle between the different regions.”

AI's Impact on Analyst Roles

30:11 to 33:58

Explore how AI is transforming the roles of financial analysts and their workflows.

“First off, is there a situation where there would be a L.A.”

Managing AI Budgets and Costs

33:59 to 37:20

Understand how companies are navigating AI budgets and the rising costs of AI tools.

“You mentioned earlier, you're trying to figure out the token approach within Pigment.”

Recruitment Challenges and Celebrity Connections

37:22 to 42:00

Hear about the difficulties of hiring key executives and a unique recruitment experience with Tom Cruise.

“and you're not allowed to say it hasn't happened yet and it's in the future because that's a cheating answer.”

Embracing Unreasonable Hospitality

42:00 to 42:36

Discover the philosophy behind creating memorable experiences for customers.

“Well, it goes with what I was telling you.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Eléonore Crespo:Spreadsheets, they cannot scale. They cannot put the right data in. They cannot communicate. You know, they cannot help different departments communicate together. So it's a massive issue for a company. And what I would see is a lot of analysts spending days, months analyzing data, reconciliating data, validating data, etc., instead of thinking about the potential trajectory of the company.

0:21Romain Nicoli:Planning for your financial future can mean wrangling a huge amount of data. Even at companies as big as Google, CFOs and strategists usually do this sort of thing in spreadsheets or old software that wasn't built for the AI era. Pigment is a French startup that says its agents can do it better. Our guest today is Pigment's co-founder and co-CEO, Eleanor Crespo. Eleanor says Pigment stands out by providing customers like Uber, Coca-Cola and Anthropic with unreasonable hospitality.

0:50Eléonore Crespo:When you come to a meeting, first of all, you know, you will bring your favorite cookies from your country. We try to create some bonds that go way beyond just, you know, the prospect is that how can we create a sort of friendship of a time? I would say whenever they have a question, they all have me on WhatsApp or on messages. And I can tell you that they send me a message, whether it's, you know, 11 p.m. on a Saturday or 7 a.m. on a Sunday. I will always answer within the next five minutes. And it's the same for the rest of my team.

1:19Romain Nicoli:Today on the podcast, we're going to talk about solving big and obvious problems, why Eleanor left a career in venture capital to start her own company, and how Pigment can help companies spend less on pricey AI tokens. Plus, why Pigment had to be built in France, where fans include President Emmanuel Macron. I'm Alex Conrad, founder and editor of Upstarts Media, and this is the Upstarts Podcast, our weekly show where we talk to emerging startup founders who are punching above their weight to take on the status quo. Eleanor, thanks for joining us on the show.

1:50Eléonore Crespo:Thank you so much for having me.

1:52Romain Nicoli:This episode is brought to you by Rippling AI, the only AI built to give you full visibility into your startup and the ability to take action across every department. So Eleanor, we're in London today, but you're usually based in the Paris area, right?

2:06Eléonore Crespo:Yeah, so I'm actually between Paris and New York now. I try to spend half of my time in the US.

2:11Romain Nicoli:So are you basically on a plane then most of the time? Most of the time. That's awesome. And what's bringing you back and forth? What exactly is Pigment doing?

2:18Eléonore Crespo:So Pigment is an AI business planning platform. So what we do is very simple. I think every CEO, every CFO right now is living through very, very complicated macroeconomic conditions where they have to react extremely fast to what's happening outside. So they have to react to, you know, supply chain shortage, tariffs, AI, etc. And so we help them make faster, smarter decisions extremely fast with our platform.

2:42Romain Nicoli:You obviously now work with some really big companies, you know, Coca-Cola and Siemens, right? And then also Anthropik as a fast-growing company. What did those big companies want from Pigment specifically that they couldn't build themselves?

2:56Eléonore Crespo:If you take these companies, so you take the Coca-Cola of this world or the Siemens or Unilever, they all have the same problems is that they have to navigate a very complex macroeconomic environment. And at the same time, they have to optimize their top line. They have to optimize their margins, their costs. They have to also be more productive and they have a big incentive to know how to use AI productively within their companies. And in order to do that, they actually need help because it's not that easy to run multiple scenarios at a time to understand how to optimize your top line, for instance.

3:27Eléonore Crespo:It's not that easy to connect your supply chain with your finance team and help them get together on the plan for the next, you know, months to know exactly what they need to do to match their supply and their demand and to connect that to the sales team and the HR team. And so these are very complex matters that they are all trying to get through. And obviously, they are all trying to automate as much as possible. Because today, obviously, with agents, you can do a lot. You can, you know, automate your analysis. You can automate modeling scenarios. You can automate decisions for you. And this is what they are looking for.

3:59Eléonore Crespo:And the idea of doing that for them is really to help them accelerate their decision making to be able to be more competitive, to be able to be more agile in this very complex environment.

4:09Romain Nicoli:For people who are learning what pigment is, obviously it works with these really big companies, but who is the ideal customer? And within that company, who is the person you're trying to get in front of to buy pigment?

4:23Eléonore Crespo:So the ideal customer can be from any industry. We serve every industry from financial services to CPG to retail. You have to be a certain size to appreciate pigment, meaning usually we start serving companies from 500 employees and more. fast-growing companies. We have a lot of the Cloud 100 to then the largest companies in the world.

4:43Romain Nicoli:And maybe upstarts in like four years or something, but not today. No, two years for you, maybe even one.

4:49Eléonore Crespo:And the reason for that is because when you put in place a tool like Pigment is because you start reaching a certain level of complexity and you need a tool to power that complexity. So that's where we start. And then the ideal profile, so usually we sell to the CFO and the VP finance, but we can also sell to your supply chain teams first or your sales operation teams first because again, the goal usually that we enter the company through a specific angle and then we expand. So with Coca-Cola, for instance, we started by the supply chain and then we started to expand into other areas of the company.

5:24Eléonore Crespo:Usually companies, you know what? They start where they have the biggest pain point. So you take ServiceNow, they had a huge pain point on workforce planning. This is where they started. And you have other customers, they will start, you know, Unilever, it was about matching their sales and their demand and their supply. So they started with SNOP. And it really depends from one customer to the next.

5:44Romain Nicoli:I want to go deeper on what the agents can do and where that goes a little bit later. But when you started this company, pre-ChatGPT, we weren't having the same discourse around AI agents. But what was the big unlock that was going to be new and powerful versus all these SaaS companies that were pitching me 10 years ago when I was running the Cloud 100 list? at Forbes?

6:06Eléonore Crespo:Right. So I started my career at Google. I'm an engineer. And what I saw at Google and that I saw as well at Index Ventures across a lot of cloud 100 companies is that whenever a CEO had to make a decision, they didn't have any tool to help them. So if you think about it, for instance, at Google, we were running pretty much all of our finance and sales workflows on spreadsheets. And spreadsheets is the most critical tool of your company. You have a problem because spreadsheets, They cannot scale. They cannot put the right data in. They cannot communicate. You know, they cannot help different departments communicate together.

6:42Eléonore Crespo:So it's a massive issue for companies. And what I would see is a lot of analysts, finance analysts, supply chain, sales ops analysts, spending days, months analyzing data, reconciliating data, validating data, etc. Instead of thinking about the potential trajectory of the company, instead of thinking strategically about how can I help the CFO of Alphabet take the best bet here. Because I'm just spending all of my time trying to reconcilia data. So for me, the big unlock was to think, I am at Google, this incredible company that is always using the best technology. Yet for this critical problem of strategic decision making, they do everything on spreadsheets.

7:20Eléonore Crespo:That's a problem.

7:21Romain Nicoli:But it's almost hard to believe because we just think, okay, there's software for every obvious function that Excel might operate in. There was, you know, ERP software. There was process automation software. Why were CFOs being forgotten here?

7:38Eléonore Crespo:That's a very good question that, you know, as a younger person, you know, just outside college, I was very surprised to see as well. But I think it's because it's not an easy problem to crack because it's mostly a database problem. It's a database problem in the sense that it's a problem of being able to model things real time. And, you know, it's like massive scale, massive amount of data. Can you imagine Coca-Cola? They use us. They have all of their SKUs on pigment. They use all of their products. They want to understand. It's a lot of products. It's a lot of countries. It's a lot of issues, butlers, everything to optimize their margins.

8:14Eléonore Crespo:It's like huge amounts of data. So that's the first thing. And they need to process that in real time. And they need to run dozens of scenarios on it. And they need to multiple their country, their business unit, their products, have the different team communicate together. And it's not an easy problem to crack. And I think that that has been the issue, is that creating that scalability while maintaining a tool that is extremely easy to use for business users had never been cracked. And this was the unlock that we found with my co-founder.

8:43Romain Nicoli:Would early career in university, Eleanor, say, wow, someday I'm going to be helping CFOs with, you know, AI agents. Like, this is my life dream. Would young Eleanor have said that or would she be surprised?

8:57Eléonore Crespo:That's a great question. Look, what I knew in university is I wanted to solve a big problem and I wanted to have a big impact. And that has always been the case. And I didn't know if it was through creating a company in that space or another space, but I knew that I wanted the freedom to create my own company. I studied entrepreneurship as part of my engineering studies because I knew I wanted to have that impact. And I do think that actually when you work on a topic that a CEO cares about or a CFO cares about, that's where you know you can have the biggest impact. I'll give you an example. Last week, I met Jamie Dimon, so the CEO of JP Morgan.

9:32Eléonore Crespo:And he asked me what, you know, what my company was doing. And when I told him, the first thing he says is, I need pigment straight away. So, you know, he put me in touch with his number two and we are starting the discussion. But it's a CEO topic. And it's a big CEO topic because everyone is struggling. Everyone is struggling with the same topics of getting to know their company in and out on what to do next in a very, very fast manner. And so that's what I love. And I do think that while I would not have told you it would have been this exact topic, I think, you know, I would not have enjoyed working on something that was not like as strategic as what we do at Pigment.

10:05Romain Nicoli:You've said in the past that you were very intense in your co-founder selection process. I think at one point you compared it like CIA level search. What was the number one thing you were looking for and how did you know when you had found the right partner?

10:19Eléonore Crespo:I knew that I was a team player and that I would probably never succeed just doing it on my own. And I think that's a personality thing that I knew I really wanted to do that, not to be at least two to create that adventure. First of all, I'm so happy that I did that choice because I think, you know, I'm on the calls with my co-founders probably three times a day. And we challenge each other every day. And I think that's what makes me a better human being every day.

10:44Romain Nicoli:I just talked to my dog three times today as a solo founder. No, but I get it. Being a solo founder is really tough. And especially if you're trying to be as ambitious as you are to build so fast. I can understand why it was helpful. Right.

10:55Eléonore Crespo:Yeah, it's tough because you have so many decisions to make every day and you need to challenge yourself and you need to be honest to yourself. So for me, it was important to have a partner. And so I knew what I was not good at and I knew what was my personality and what I needed to potentially create something massive. And so I actually did a list of the people that I admired the most. and I knew that, you know, if I really wanted to have an impact in the enterprise, what we would need, first of all, was the best engineering team in the world. So I needed a co-founder that could help me put that together.

11:29Eléonore Crespo:And Romain, my co-founder, happened to be the co-founder of a company called Criteo that was probably one of the very first, if not the first French unicorn that IPO'd on the NASDAQ in 2013. And he had the best reputation in the market. We knew he was one of the best European, if not world-class CTOs that had put together one of the best engineering team. So for me, it was a dream. It was a dream. So I had done my list and I started to ask people. So I did a big due diligence about Romain. I knew so many, you know, we had so many people in common and, you know, everybody was giving me the most glowing feedback.

12:04Eléonore Crespo:So indeed, I did my own CIA due diligence and he had no idea. And I went to him. and you know honestly it was a dream for me to work with him because I had zero track record and you know he could have worked with literally anyone he wanted to given what he had built and I went to him and I literally prepared a book of ideas I had written more than probably about a hundred pages memo about these are the things I would like to do with you what do you think and so we started to spend an afternoon on another afternoon on another afternoon and he was like I don't know let's go and I remember because I told my grandmother it was at Christmas time I said I think I found my co-founder and I think we're going to do something big.

12:44Eléonore Crespo:And she was like, I trust you, Eleanor. And that's how it started. So, yeah. And really, you know, I have to say it's an advice I would give to literally anyone. Take your time to find a founder if you want one, because that will make or break your company. And I would have never been able to build even a tenth of pigment without Romain. That's for sure.

13:04Romain Nicoli:This book of ideas, were they all within the idea that the broad area that became Pigment, or were they different types of startups, and then you guys zoomed in on Pigment over time?

13:14Eléonore Crespo:They were all B2B enterprise ideas that were more or less close to what we do at Pigment, but tackling the problem through different angles and trying to think through my VC lens about what could create a massive company. You know, how big is the time, the market? How fast can we go after the opportunity? What does a competitive landscape look like? What does the user look like, etc.? So it was very complete. And I think today, obviously, with AI, you can do that extremely fast, too. So I think finding the idea is not anymore the bottleneck. But at the time, you know, it required quite some effort to really think about uniterate on the idea.

13:51Eléonore Crespo:So we were glad that we had done that.

13:55Romain Nicoli:Nowadays, in this moment, especially AI founders are trying to vibe code a product really fast, maybe even raise money before they have a product. Everything is like very fast. You know, let's do something in the matter of days. Was there any temptation to be doing that? Or did you feel all along, hey, we need to be really careful. We need to take our time. And I'm curious if like now you're doing it all over again, would it have been hard to take that approach? You know?

14:18Eléonore Crespo:Yeah, it's a very good question. So first of all, I think even at that time, we were doing a lot of prototypes because you need to show to customers, to design partners what it looks like. And it's the worst thing you can do as a founder if you stay within your room and don't try it out with people outside as fast as you can. So for us, you know, day one, we were already showing it to people, to potential customers that actually became customers all the time to say, do you think we are doing it right? Now, clearly, you can go even faster with AI. But I think in our play, in what we do, you have to understand that we are a true operating systems for companies.

Read the full transcript

14:51Eléonore Crespo:We are true infrastructure play. And it's not a play that you build in minutes. It's something that actually took us like a few years to put together. If you think about it, I would compare it to someone that wants to build a Ferrari versus tuning a car. You know, it's like I think everybody today can take some YouTube tutorials and tune a car and you can find every piece to make your own car if you want to. But once you really want to build a very complicated engine, very complex, very well designed, that will last forever, that can go extremely fast, it's another play. And the Ferrari designer will not take, you know, two days to build this Ferrari.

15:25Eléonore Crespo:It will probably require a few years to put it together. And that's a bit of story. So I think while we are able to do prototypes and while we were able to do prototypes, still when you want to do a true infrastructure play, even today, it's not a play, you know, if you want to build Databricks today, it's not going to take you two days. I can tell you that. Now, you know, starting to get the idea of Databricks can get you done in a few days, but not the full story of what you want to build. We play a lot with what we call unreasonable hospitality as a concept, where we really try to be memorable with our customers.

15:55Eléonore Crespo:You know, when we come to meetings, the way we show up, the way we deliver the experience when they buy pigment, etc., we make it extremely memorable. And we spend a lot of time trying to really deliver delight there.

16:08Romain Nicoli:Okay, I got to know an example of unreasonable hospitality. What does that look like?

16:12Eléonore Crespo:And it's the same, by the way, for customers or also for candidates. So I can give you examples on both sides. So, you know, for customers, it means, well, when you come to a meeting, first of all, you know, you will bring your favorite cookies from your country, perhaps even with the design of the company. The other day I was at Meta. And so, you know, we brought some cookies with Meta and Pigment and we invite them to an extraordinary event that we do with the team. And we try to create some bonds that go way beyond just, you know, the prospect is that how can we, you know, create a sort of friendship of a time, I would say.

16:43Eléonore Crespo:we make sure that we deliver fantastic experience when they become customers meaning that you know whenever they have a question they all have me on WhatsApp or on messages and I can tell you that they send me a message whether it's you know 11pm on a Saturday or you know 7am on a Sunday I will always answer within the next 5 minutes and it's the same for the rest of my team so we are extremely reactive we try to fix issues we never push back on issues we really try to be there all the time and we always do the extra mile. If a customer needs us for a specific problem, we'd fly to meet them. It's really across the board.

17:19Eléonore Crespo:Like think about the experience you would have when you buy a very luxury car again or a luxury house. We try to reproduce the same. Or when you go to a palace, we try to reproduce the same internally.

17:28Romain Nicoli:But Ferrari is not mass produced at a scale that everyone drives a Ferrari.

17:33Eléonore Crespo:Yeah.

17:33Romain Nicoli:How do you scale that without burning out or just literally not having enough time in the day to respond to every message, especially as you add more customers?

17:41Eléonore Crespo:Right. Right. Well, I think, you know, first it's like we serve the enterprise. So we have the ability. It's not that we serve, you know, we don't have like 50K customers. It's not a must play here. It's really a play where we serve the largest companies in the world with a solution that is, you know, tailored for them. So we have this ability, first of all, to be more wide-loved in the treatment. And it makes a whole difference. And this is why they love us. They don't just love the product. They love the experience. So that's one thing. And second is about finding a way to scale that experience, too.

18:11Eléonore Crespo:So, you know, how do you automate ways for people to get served extremely fast, you know, when they have an issue and using AI to do that, making sure that, you know, it's not only me on text message, but it's my entire leadership team and the people that are, you know, reporting to my leaders. So making sure that this philosophy spreads out across the company at every level for every single individual. It's extremely clear that when you are at Pigment, it's like if you work in the best hotel in the world, that's the experience you need to deliver to a customer.

18:39Romain Nicoli:How did you both get excited about being co-CEOs and how did you make that work in practice? I know you've talked a lot about how it's great and you divide and conquer, the engineering, the sales. But in terms of actually getting the relationship on even footing, what was the biggest lesson or how did you really set that up?

18:56Eléonore Crespo:First of all, I think having a mix in any company of up-and-comer like I was and experienced people is something that I would recommend everybody. And this is what we have in the C-Level team. And I do think that having some people that have true company expertise, that have done it before, with, you know, people that have the potential, the energy and the go for it attitude has been fantastic for us. So I think, first of all, this complement is great and it's been a strength for us. And I think Romain is really the definition of humility and is a true team player. And I think he knows and he believes in potential.

19:38Eléonore Crespo:And I think there was never even a question of him, you know, having more equity than me or him like having more leadership than me. Like from day one, he was like put each other on equal foot, you know, and he would have told me differently. I would probably have accepted because I would have understood where he was coming from. But I think that's also the key of trying to find a founder that has very, very low ego because that's for me a big recipe for success. so I think I was extremely lucky to have someone like him on my side and I have to say that what has been great is that for me he's been a true mentor you know I'm learning every day he has done so many mistakes with Criteo so many successes that you know we can take sometimes a fast pass and I think I bring him something completely different which is you know sometimes I have no clue and I come with my own freshness and it actually brings a lot of value because things have been done a certain way maybe for 20 years but even more now than ever in the time of AI perhaps we should do them completely differently and it will work even better.

20:37Romain Nicoli:I mean, we're seeing a few co-CEO successes in progress right now, but historically, at least when I started my career covering startups, it was considered a real challenge to make that work, in part because of the idea of there's a crisis. Who has the final say? Like, do you guys need to reach consensus on every decision?

20:56Eléonore Crespo:So it's interesting. We never had a single conflict about who has the final say. So it probably means that we are quite well organized around that. What we've tried to do is to define extremely well our roles and responsibilities. We know exactly, you know, he leads tech, product, HR. I lead everything else. And we are extremely clear about that. And, you know, when it comes to taking a decision, I think we've worked together for long enough to know exactly when we need each other's opinion or not. So, you know, there are a lot of tactical things that he will happily do without me that I will happily do without him.

21:30Eléonore Crespo:And then when it comes to a very big company milestone where we have to decide about the product direction, well, we do it together and we do a two hours workshop together to decide what to do. So I think as long as your roles are extremely well defined, not only to you, but to the rest of the company, then people know exactly where to go. We have this concept internally that is a concept of direct ownership, DRI, like making sure that there is always a single individual that is in honor of a specific topic. and that is something that we should do on everything. I think it's really, really important and this is what we do really well with Romain.

22:06Eléonore Crespo:It's like everybody knows what I'm responsible about, people know what he's responsible about and so it makes it very, very straightforward. I have to say it works very well, very fluid but it's also, I think, a question of we have also, I think, the right personalities to make it work so that it never becomes a conflict where it could be.

22:22Romain Nicoli:Got it. But it sounds like if you don't have that clear framework and definitions of where someone feels confident that they have, you know, the credibility, that's where people could step on each other's toes. Like, it sounds like a lot of communication is needed up front to make that work.

22:40Eléonore Crespo:Absolutely. And I think you have to have the right EQ to also understand when, you know, the person might get upset if you don't ask them for their own opinion, right? So it's, but I think it becomes quite obvious because, you know, it's, there are decisions that truly matter for the company and there are some that, you know, are not going to move the needle. Like one of the last example, this weekend I went to SF to actually close a C-level candidate for pigment. And, you know, for me, this is a super critical element where I'm going to call, you know, my co-founder probably two, three times so that we agree on, is that the right person?

23:15Eléonore Crespo:What are we really trying to judge here? What is the right package to make them come, et cetera, et cetera. So this is a critical decision for the company. Whereas, you know, there are dozens of other decisions that I'm going to make on my own, like, you know, during the day.

23:29Romain Nicoli:You've raised a few hundred million in capital. Does Romain let you pick the VCs because you're a former VC or does he care about that? Oh, yeah.

23:37Eléonore Crespo:That's the type of thing he doesn't care about, you know, and he's going to trust me. I knew it.

23:41Romain Nicoli:I knew that would be one for you for sure. Was there any temptation for you guys to move to the Bay Area at some point? I mean, obviously, now there's a lot of question about, you know, should AI startups all be there? You guys are AI enabled. I mean, and if it was clear to stay in Paris, why and how has that helped?

24:00Eléonore Crespo:Right. So, well, look, first of all, it's true that I'm spending a lot of time in the Bay Area. We have 60 percent of our revenue in the U.S. and we serve some of the largest tech companies out there from Palo Alto Networks to ServiceNow to, you know, like all of the Cloud 100 like Notion or Figma. So for me, I've spent actually a lot of time there. So sometimes I'm like, maybe I should just move. But look, being in Paris is an incredible edge for us. So I was telling you, my co-founder is fabulous. One of the things he has done is creating an engineering team that has been probably one of the most loyal team in tech history.

24:35Eléonore Crespo:Meaning that since we started Pigment, we had zero regrettable attrition in the engineering team. Can you believe that? You know, seven years in the world. That's as good as it gets. So, and it's just like, that's the reality that in France, we have some of the best engineers in the world. We have some of the best AI schools in the world. So we know what we're doing. And people are loyal. They believe in the mission. They, you know, they love what we do at Pigment. We have an amazing, amazing, amazing team of the smartest talent. And that's something that would have been impossible for me to replicate in the Bay Area.

25:08Eléonore Crespo:I remember having a discussion the other day with one of my friends who is the founder of Intercom. and that's what's telling me when you're in the barrier if you're not anthropic or open AI today it's like when you go to a Chinese buffet and you have lobster and you have whatever let's say a good steak and then you have the noodle and then you have the rice and so you come and anthropic and open AI they take the lobster and the good steak and then you come and there are some noodles left and you're like oh my god that's what happens today and it's very very very hard not only to compete but to retain the talent so I think for us building in Europe has been an incredible strength it has also helped us think globally day one multicultural we have a new global head of operation she told me she's astonished by the level of multicultural element that we have at Pigment we have people from every country we are super proud of that and I think sometimes it's harder to replicate in the Bay Area where you are more in a sort of different mindset Now, clearly, I don't think my head of product, my co-founder, myself, we spend so much time there.

26:13Eléonore Crespo:Like, clearly, a lot of the innovation is happening there. We are partnered with Entropic. We are partnered with OpenAI. We spend a ton of time with our teams. And we will keep doing so for the foreseeable future because it's true that that's where a lot of things are happening today.

26:26Romain Nicoli:So you team up with them so that they'll let you in the buffet as well? Exactly. Yeah, yeah. Fair enough. I do find it interesting that in the conversations I see you have at conferences like London Tech Week or Human X or some of these shows, you know, you have stressed that you sell a lot to the U.S. market. You are this global business. At the same time, your countrymen and women are very proud of you in France. Like Emmanuel Macron has like praised your last funding round, called you a unicorn, whether you were ready to say that or not. And I feel like Pigment, Mistral, there's a couple of companies that get people really excited.

27:02Romain Nicoli:Is there a way for you to lean into sort of the excitement and sort of the desire to have these success stories in France while also leveraging, you know, the talent in the U.S. and sort of being that global business at the same time?

27:15Eléonore Crespo:Yeah, it's a hard one because first, we are so lucky, so lucky in France to have the support of the government. I cannot thank Emmanuel Macron enough for all the efforts he's putting to actually put light on what we do. And I think it does that. I don't think there is a lot of government that does that as well as he does it in France. So, you know, extremely lucky. We, you know, thanks to the government, have met so many of the top CEOs in France that are now buying pigment. We, you know, we've been able to abuse here, leverage a lot around, you know, the sovereign AI that mystery is trying to create, etc.

27:55Eléonore Crespo:So we have been very lucky in that regard. But at the same time, it's true that, you know, I'm here to create, you know, a global success from France. And so for me, it's really important to juggle between the different regions. But it's true that it's been a fantastic upstart, actually, to what we're doing. And I think I would love to see more governments doing exactly what's happening because it's not easy when you start a company, especially with us, you know, selling to enterprise in Europe where companies are super conservative. You know, we have customers now like Essencef or Siemens that, you know, are not like potentially the companies that would buy pigment first.

28:37Eléonore Crespo:And I think he really managed really, really well to make sure that small companies would get considered and recognized.

28:45Romain Nicoli:When the president of your country goes against the press release and calls you a billion dollar startup, do you just let it go? Do you just laugh? What is the reaction?

28:54Eléonore Crespo:It was like so beat, you know? Yeah, I think we didn't want to announce anything because we found that it was not necessarily the milestone we wanted to celebrate. But yeah, I think, you know.

29:05Romain Nicoli:I think it's great because now if anyone looks you up on, you know, ChatGPT or Claude or Google and they say, what's the valuation? It will say, source, president of France.

29:15Eléonore Crespo:I mean, that's pretty cool.

29:17Romain Nicoli:These days, you can chat with AI about almost any business problems. Rippling AI is built to actually solve them. That's because Rippling AI is built on your live workforce data. That gives you full visibility into your startup and the ability to take action across every department. Say you have a new hire starting Monday. Just ask Rippling AI to get them set up and it will provision their laptop, grant them access to your systems, assign their onboarding tasks, and loop in their manager. But it doesn't stop there. Rippling AI can flag missing paperwork, set up their first one-on-ones, and make sure they're ready to ship on day one.

29:51Romain Nicoli:All you have to do is tap confirm and get back to building. Don't settle for AI that's all talk. Head to rippling.ai slash upstarts and get AI that turns insights into action. That's R-I-P-P-L-I-N-G dot AI slash upstarts.

30:08Eléonore Crespo:Sign up for exclusive access today.

30:11Romain Nicoli:Okay, well, let's talk about how AI challenges that and maybe unlocks more there. First off, is there a situation where there would be a L.A. and or agent that might be responding to some of these people? Or do you think it's important that it be a human?

30:25Eléonore Crespo:So actually, we use already some agents to do my emails. You know, we try to really automate as much as possible around the way, you know, to make sure that I respond extremely fast to customers, etc. So we will always do the last mile. But, you know, having all the pre-draft, making sure that it's really like simple things. But you have a meeting straight after. Within the next 15 minutes, I want to have reply to that customer with all the data points that we discuss during the meetings and the follow-ups and the next steps. So, you know, we use AI to record the meeting. We use AI to draft the email.

30:57Eléonore Crespo:And then I just have to send it. And it's very easy. Otherwise, it would have taken me probably 20 minutes to do that. Now it's just a click. So there is a lot you can automate, obviously, around that, but around many other processes to make sure that you are extremely responsive for support. You know, we've trained agents on all of the issues that our customers face, even the most complicated one, so that the first answer is drafted in a wonderful way. And I can tell you that our support is not easy. Like, you know, when you have some issues within pigment, they are usually complex to solve. But we have so much history and knowledge that, you know, again, we can really go to the last mile.

31:33Eléonore Crespo:So we try to do that on every single interaction with the customer. So we cannot automate yet doing the cookies for you, but you can automate a lot around the way you interact and work with customers.

31:45Romain Nicoli:On the customer side, at the same time, there's these AI agents that the CFO might be using or their team is using. What is the conversation like with the employees at these companies who might say, okay, well, this is great for the CFO, but maybe as an analyst, I'm going to be unnecessary because AI will do most of the job or we don't need as many analysts as before, which obviously a lot of companies say, we're not going to decrease our headcount because of AI. But if you are a public company, especially a huge one, and these tools are saving you a lot of money, that's always somewhere in your mind.

32:18Eléonore Crespo:It's interesting because our analysts are the primary users of Pigment and they are actually super happy that we managed to automate part of what they do because we automate the cumbersome parts of their jobs. So, you know, when you're an analyst and you are being under pressure, the CFO is asking you to prepare the board deck and you have you know a week to do it and you know you're going to spend your entire weekend and evenings on it because there is just so much work to be done well you are very happy if I come and tell you I'm going to help you automate and do a pre-draft of the analysis I'm going to help you build that model that's going to give you the answer to that specific question that the CFO asks you so they're actually very happy because most finance teams, whenever I show them what we do, they tell me they spend 80 % of their time doing data reconciliation, data cleaning, trying to understand, to make sense of the data, finding data across teams, you know, gathering data from the different shareholders that they have to work with.

33:17Eléonore Crespo:That's not very interesting. And they want to spend more time on the 20%, but they are so stretched. You know, it's very rare that I meet a CFO and he tells me my finance team has nothing to do. It's like, no, they have so much to do and they are under so much pressure that everything you can do to decrease that 80 % to have them focus on strategy matters is critical. So, you know, probably there are companies that are going to reduce headcount. Probably there are companies that are going to have to, you know, rethink their processes, et cetera, with AI, that's for sure. And probably there are finance teams that are too big today for the world we live in.

33:50Eléonore Crespo:But I can tell you that for most of our customers, well, they actually need AI to help them be more intelligent. Because today, the problem is that they don't have enough time to spend on strategy matters.

34:01Romain Nicoli:You mentioned earlier, you're trying to figure out the token approach within Pigment. Is this also something coming up with customers in terms of sort of budget for new generation software tools? We obviously have seen some discourse about AI tokens in general, people blowing through budgets a lot faster than they expected. So I'm curious, A, what's the latest inside Pigment? And then B, kind of what's the sentiment right now you're hearing from these big companies that you work with?

34:29Eléonore Crespo:So we already have some customers using us for token management. Datadog is an example that has built an app on Pigment to help them with that. And now we want to offer that to literally every customer. A customer of ours is Uber. They are not using us for token management, but you probably heard their CTO that has already exploded their budget. So I want to see if we can help them on that.

34:49Romain Nicoli:He set off this whole news cycle by saying what, after a quarter and a half, they had spent a year's worth of resources on tokens. So he would be your dream customer. Yes. Yeah. Yeah.

35:00Eléonore Crespo:And there are customers on other sides of the business. So I would love to build something with them now. Well, it's going to be a big topic for everyone. And it's a topic that goes beyond just understanding your cost. It's also about what models do you want to use? You know, for us, it's a question that we're asking ourselves today. we are doing a partnership right now with Fireworks to work on open source models for certain part of what we do and using more expensive models for other parts. And I think that's really the reality we live in today is how to optimize your cost because you can't just build everything on very, very expensive models anymore.

35:38Eléonore Crespo:We really want the insight that I see that I am on so many customer meetings and every CFO is like, you know, that's completely getting out of control. and the problem is that our CEO is pressuring us for everybody in the company to use AI but we don't have any right way to think about it. Like we want to be AI peeled but we don't really know like where we should put the limits, et cetera. So what I'm building right now, the offer is very simple. It's an offer where you can obviously compare the costs across your model but also connect that to IT and connect that to different departments and start giving budgets to different departments based on the needs and on what they need to do, et cetera and seeing that completely holistically across vendors.

36:18Eléonore Crespo:So, yeah, the idea would be to start putting a bit of control around that.

36:21Romain Nicoli:So you say to the CFO, pay for our AI tools so you don't pay as much for everyone else's AI tools?

36:26Eléonore Crespo:Absolutely, of course.

36:27Romain Nicoli:And then does Anthropik get mad at you as a partner when you do that? Or do they understand this is the way things have to go?

36:33Eléonore Crespo:I mean, they have to go this way anyway. You know, I don't think anyone will be happy if they have committed$10 million to a frontier lab. And after two months, they have already exploded the budget. You know, it's not going to. so we need to find sustainable way to make that happen and i think it's not that i think every company is happy to spend more on ai it's more like controlling the cost and knowing what you are you know where and why you are investing in certain areas but i do think that the budget will keep growing exponentially anyway so i do believe that you know anthropic and open ai have the brightest days in front of them so that's not so much the issue the issue is more being quite intentional because I feel like most of our customers have done experiment mode for the past two years and now they really want to be intentional about where it matters, where it makes sense and how to spend.

37:21Romain Nicoli:When you think about the hardest period of the company so far and you're not allowed to say it hasn't happened yet and it's in the future because that's a cheating answer. What has been the moment where you felt the most like an upstart where you were punching above your weight or sort of, you know, back to the wall?

37:35Eléonore Crespo:Right. I have to say my answer would have been the contrary. Every day is hard. Every day is hard. I've heard that one before. It's banned from the show.

37:42Romain Nicoli:You have to come up with a different one.

37:43Eléonore Crespo:No, no, no, no, absolutely. No, there are multiple examples. So, well, one of them that I think was hard, but at the same time, absolutely exciting was to be the first to market to launch our agents. You know, when the technology was just nascent, it was, you know, making the bet to say, we know that it's going to be the future of the company and we have to go really, really hard at that. and it's not easy because you know it's a new technology the company obviously we are an infrastructure company so we were focused on building our engine and and you know making the company as scalable as possible with our engine it's always a question of focus you know every company is trying to be as focused as possible so when you have this new technology and you know it's going to be the future and that's where you can make a revolution you know bringing everybody with you uh is is quite something and it was totally worth it and And it's been incredible.

38:35Eléonore Crespo:Like the journey since we made that choice, you know, about like a year and a half ago, two years ago, has been phenomenal. But, you know, it's one of these moments where, you know, you are at the beginning of a revolution and you just can't miss it. So I think that that has been a big one. I have another big one, if you want.

38:52Romain Nicoli:Of course. I'm not going to say no. No, give me more. We're at the buffet.

38:57Eléonore Crespo:So speaking of buffet. So what about candidates? For me, I think that has been one where I have put my back against the wall many times since I started Pigment. The issue with a lot of fast-growing companies is to have the executive team scale with you. And it's very hard because, you know, people sometimes, you know, don't feel that they have the energy to keep going. Or sometimes, you know, the company outpaces them and you have to realize that very early. And for us, you know, last year I had to find my CRO, so my chief revenue officer. And I have to say that has been one of the hardest moments for me of Pigment because having the right chief revenue officer can truly completely make or break your company in enterprise B2B.

39:40Eléonore Crespo:And you make a mistake. If you make a mistake, you realize a year later that you've made a mistake and potentially your company is out of business by that time. So for me, it has been a quest. But when I say a quest, it has really been a quest. it's been nine months of mapping of intense research of understanding what are the best schools in the world for CROs who can be the best candidates doing calibration meeting like dozens of people that were supposed to be the best CROs in the world for them to give me advice etc and it has been one of the hardest moments because at some point I thought I would never find the right person for pigment that accepts to come at our stage with the right potential with the right growth curve ahead of them that would be able to scale the company through multiple billions of revenue.

40:28Eléonore Crespo:And I found my candidate. I was very proud. I closed him by having him meet Tom Cruise at the time. That was a big one, thanks to Iconic, one of our investors. You got to tell us how that happened.

40:39Romain Nicoli:How that happened.

40:41Eléonore Crespo:So Iconic, I don't know if you know them well, but Divesh, the founder of Iconic, is also at the head of the family office of Iconic. What's great about Iconic is that they have many tech co-founders as part of their family office, but also obviously it's quite fun. They have a lot of celebrities. And so Tom Cruise is one of their friends and he is spending a lot of time with founders at Iconic. I know he spent some time with my friend at Eleven Labs, for instance, you know, working on some voice matters, etc. and Tom Cruise a good friend of Divesh was actually coming to the Iconic office for a very you know private meeting where I got invited and the idea was to understand from him you know how he thinks about filmmaking and you know what it takes to be such a you know such a great actor over the years and everything else and so we had an hour and a half with him and I begged Divesh and Iconic I said can I bring my candidate because I think you know that would make a big difference and I said yeah let's go and so So, you know, I think it was quite, you know, quite a lifetime moment for him and for myself because Tom Cruise, you know, he's, I mean, understanding, you know, his life is absolutely inspiring.

41:52Eléonore Crespo:It's really crazy the type of thing he's done in his life. And so that's how we managed to get him through there.

41:58Romain Nicoli:So any founder watching this show now has to introduce their next candidate to someone as cool as Tom Cruise, basically. You've just raised the bar quite a lot.

42:07Eléonore Crespo:Well, it goes with what I was telling you. that's what I call unreasonable hospitality. So, you know, I'm bringing some customers to meet Céline Dion in two months in Paris. And, you know, that's the type of thing we try to do. And I think that's what makes life fun. Yeah, I think I want to enjoy every day of the journey and I want my customers to do as well.

42:26Romain Nicoli:And are you enjoying it right now? It seems like it.

42:29Eléonore Crespo:Yeah, no, it's a journey of my life.

42:32Romain Nicoli:Fantastic. Well, thank you so much for coming on the show.

42:35Eléonore Crespo:Thank you. Thank you so much for having me. Thank you.

From the publisher

The world's big companies are blowing through their AI budgets. At French startup unicorn Pigment, Eléonore Crespo wants to help.

"Our CEO is pressuring us for everybody in the company to use AI, but we don't have any right way to think about it," Crespo says she hears chief financial officers complain. “It’s going to be a big topic for everyone, and it goes beyond just understanding your costs."

Pigment, founded by co-CEOs Crespo and Criteo veteran Romain Nicoli in 2019, fights runaway AI spending with AI. Its agents help CFOs and finance teams at customers like Anthropic, Notion, Coca-Cola and Unilever make faster, better financial decisions.

For analysts, Crespo argues her tools can free up much of the 80% of time they currently spend sifting through data; her (perhaps optimistic) hope is that they can then focus on higher-level decisions, without fearing for their jobs.

On The Upstarts Podcast, Crespo talks about how building in Paris helps Pigment compete; how a culture of “unreasonable hospitality” makes it win; and how careful communication makes a co-CEO setup work.

Plus, she shares her Upstart Moment: landing a key executive hire, with an in-person assist from the actor Tom Cruise.

Chapters:

00:00 Introduction
02:18 What Pigment does, and who it helps
06:06 Seeing Google’s spreadsheet problem first-hand
10:05 Finding a co-founder, CIA style, and a product, Ferrari style
15:49 Cookies, WhatsApp and “unreasonable hospitality” as strategy
18:56 How to make startup co-CEOS work
24:00 Building in Paris versus the Bay Area’s “talent buffet”
30:11 How AI agents change CFO and analyst jobs
34:05 Bringing AI token spending under control
39:02 How Tom Cruise helped close Pigment’s CRO hire

Eléonore's LinkedIn

Pigment

For more, visit https://www.upstartsmedia.com/

Season 2 of the Upstarts Podcast is presented by ⁠⁠⁠Rippling⁠⁠

Produced & edited by Eric Johnson from LightningPod

More from The Upstarts Podcast

All 23 episodes
Pigment’s Eléonore Crespo: Building France’s Next Unicorn With AI Agents And Tom CruiseThe Upstarts Podcast · 43 min
Listen in VO