S13E5: From Palantir to Hex — How Barry McCardel and Caitlin Colgrove Built the Modern Data Workspace

redo your whole pitch deck just selling the problem. Be like, I understand this problem in and out. It's a huge problem. No one's solved it yet. I totally get it. And what's interesting is I think that is actually what winds up attracting early investment more than your perfectly formed idea. Because even if you don't realize it, the VC on the other side, they know that things change a lot. People are attracted to problems. And I think if you try to explain your whole idea and your master plan, maybe they'll buy it.

Welcome back to Season 13 of The Room Podcast. Claudia, can you believe we've been doing this this long?

No, it's been 5 years. So many episodes, so many incredible conversations telling the stories of some of the world's most iconic founders and funders. And whether our listeners are new for the first time today or have been with us from Season 1, we're so grateful for the community we've built.

Completely. Uh, we've had the mission to open the door to the room where it happens for those entire 5 years and have been able to do so over 120 episodes with hard-earned lessons from some of the top technology and consumer founders of our time.

And we've been through a lot personally as well.

We have indeed.

I built, scaled, sold my startup Priv. You were our first investor, so you've been there through the entire journey. And you've also recently graduated from HBS. You're now building at Vercel. And through it all, we've brought the funder and founder perspective.

Yeah, it's really been an incredible journey, and I'm thrilled that I'm back full-time in San Francisco, and we're probably here to stay. Uh, that means that there's more for us in terms of growth, both inside and outside of the physical room. But what's important to have really out there is these stories and We've been able to do so over incredible guests from the CEO and founders of Flexport, Perplexity, Particle, Zapier, Zillow, Vanta. I could go on. These guests and their enterprises have generated over $100 billion enterprise value and really, frankly, moved the markets.

And growing, especially in 2025 with everything that's going on in AI. So I'm so excited to share Season 13 stories that are maybe a little bit more AI-focused than past seasons.

Absolutely. And you can find us IRL and URL, especially with our events such as our annual conference, Inside Summit, or upcoming SF Tech Week events and beyond. Beyond.

And if you wanna get information around those events, subscribe to The Room Podcast and subscribe to our newsletter at theroompodcast.com.

Perkins Coie supports the most innovative entrepreneurs and investors in fast-moving and high-growth sectors, addressing their myriad of legal needs. But the firm doesn't just provide end-to-end legal and business counseling to its startup clients. It also facilitates introductions to key advisors and sources of capital. Perkins Coie's interactive website, Startup Percolator, offers access to programs, resources, and rich dynamic content designed to assist entrepreneurs on their startup journey. To learn more, go to startuppercolator.com and perkinscoie.com.

This podcast is brought to you by Mercury, the banking platform businesses like The Room Podcast use to simplify their finances. It's time banking did more than hold your money. Now it can. With Mercury, you can pay bills in seconds, close the books faster, and even send invoices. Not only does Mercury do away with a patchwork of tools, it eliminates guesswork, giving you complete and accurate visibility into your business's finances, all from one account. Apply in minutes at mercury.com.

A few short weeks ago, we dove into the future of data and AI with Barry McArdle and Caitlin Colgrove, the co-founders of Hex, at our live podcast recording during SF Tech Week. For those who haven't seen their bus ads around San Francisco, Hex is the AI-powered analytics workspace transforming how teams explore data, share insights, and just make better decisions. Since launching in 2019, Hex has become one of the fastest-growing platforms in modern analytics at companies like Calendly, Ramp, and even Vercel. Barry, CEO, and Caitlin, CTO, both began their careers at Palantir after graduating from Northwestern and Stanford Retrospectively, Barry went on to lead product and deployment teams at Palantir and later helped to build the clinical trial platform TrialSpark. Caitlin spent her career designing data tools that empower people during her time at Palantir and then to Remix, where she worked on city-scale transit analytics. Their mission today: to make everyone a data person. In this episode, Barry and Caitlin opened up about their co-founding a company grounded in their deep trust from having worked together for years. We talked about a topic that is very important but rarely explored, like navigating board dynamics and working to have your board work for you in sometimes environments where the incentives can feel misaligned. Ultimately, we also talked about their decision to reintegrate their AI team, which they had spun out with HexMagic, to bring it back in. This conversation felt particularly prescient as we're all navigating how to scale AI products that last. Of course, we close with some hot takes about what data will hold in the future as well. Let's get started and open the door. Hi everyone. We're so excited to have you all. Thank you so much for being here. Really, this is so special for Claudia and myself. We've been building The Room podcast for 5 years. We're so grateful for our sustained partnerships with Perkins as well as Mercury, and we've been doing Tech Week events since 2022. Fun fact, I don't think there's anyone from Andreessen here yet. Andreessen did not invent SF Tech Week. Actually, people in LA invented LA Tech Week, and Andreessen decided to pick it up and say it was theirs. So we love Andreessen for that. And their original one was in 2022.

They're good venture investors.

They're really good venture investors.

Someone has an idea, they see it, they're like, that's a good idea.

Yeah, exactly. So we're delighted. This is our fourth SF Tech Week. We love to do these live recordings as such. I'm sure people will trickle in. So with that, we'll go ahead and get started. Today we have an incredibly fun story with the Hex co-founders. For those of you who don't know Hex, it is the world's most popular platform for collaborative data analysis. We have Barry and Caitlin here. Barry is the CEO. Prior to Hex, he led development and deployment for product teams at Palantir, working on solutions to data problems all over the world. He was then at TrialSpark Information, where he helped to stand up their clinical trial platform. He has a BS from Northwestern University where he focused on network science. Not really sure what that means, but I love that. And then of course, Caitlin, she is the co-founder of Hex. She spent her career building data tools and analytics, first at Palantir, then at Remix, and co-founded Hex to really make the vision of everyone a data person, which I love that mindset. She focuses on their product and technology ecosystem, especially in the era of AI. All right, let's jump into it. But before we dive into the founding story of Hex itself, We'd love to start at the beginning for both of you. Where did you grow up and how has that shaped your view of the world? Kaitlyn, do you want to get us started?

Sure. Technically, I was born in San Francisco, but when I was about 4 weeks old, we moved to Boston and I spent most of my early years there. Maybe the most formative thing is that I'm still a Red Sox fan. I spent a lot— very early, I actually knew that I wanted to do something related to software engineering. I taught myself Python when I was in middle school, and it became pretty clear that was the thing that I liked doing in my spare time. I just like writing code. Code. And along with that, I also had my eyes set on Stanford for a really long time. And so I finally came back out here as soon as I could, and I asked my parents, like, why did you ever leave the place with the most beautiful weather in the world all year round? I've been here since then with Palantir Remix.

And then the Stanford name drop wasn't in the bio your team gave me, just so you know.

Barry, what about you? I was technically born in Delaware, just like Hex.

Yeah, exactly.

Yeah, just like every C-Corp. Yes. I actually lived there for 3 days. My dad was there on a rotation during his residency and they thought they were— anyway, it's a long story. Anyway, I grew up in Michigan, in Lansing, Michigan. I, anyone here been to Lansing, Michigan? You have? I don't recommend visiting. It, it's fine. Not to slag Lansing, it's just not a lot going on. But you know, very interesting community. It was like very Midwestern, very down to earth. There's a university in East Lansing, Michigan State. So actually a lot of like my friends had parents that were like professors and scientists and it was actually really cool. I had like a group of nerds I hung out with a lot growing up and wound up teaching myself to code. And I was also very creative. I did like a lot of film stuff in high school. And so I actually went to college at Northwestern having no idea what I wanted to do. I was like, I'm gonna do econ or film, which is a very two different things. Yeah. I think I wound up stumbling into network science, which is, I think data science wasn't really a term yet. This is like 2010, 2011. And social network analysis was this sort of budding practice, which was getting hot because social networks were getting hot. But the analytical techniques actually have application far beyond what you would think of as like a social network, like Facebook or X or whatever. I took this class and really got enamored by it. And the data analysis work as part of it was like super cool and was writing like this code in R that was like, wow, what's going on with this? And I convinced the professor to let me be like in the research lab as an undergrad researcher. And That just plunged me into like, this is a whole thing. This whole data is like a whole thing you can do. That was really cool. So it was very formative.

Did you always think you were going to become a founder?

No, uh, heck, this kind of dovetails into how we started the company, but both of us, I think we're kind of reluctant founders. For me, the company was very much a product of like personal frustration with the state of data tools. And I started as more of like a buyer than a builder. I was actually just trying to solve a problem we had at the company I was at. And yeah, it just dawned on me one day, like, why hasn't anyone built something here? And of course your first instinct is sometimes maybe that's for good reason. But like the more I spent time with it, it was like, there's obviously something could work here. I was very fortunate to have a couple amazing people that I had worked with at Palantir, Glenn, who's our third co-founder, and Caitlin. And I was fortunate enough that I could convince them that we should all take a crack at this. And a few years later, we're certainly beyond where I would've ever expected to be.

Caitlin, you guys were at Palantir, you all met. How did that experience shape your conviction in starting something? There's been a number of Palantir founders that have spun out and started something. And so I'm just so curious, like, what was in the water there?

There's a couple of things, both with our specific experience and then also a lot of things that Palantir, the company, got right that I think has produced a lot of founders at the other side. I'll start with the Palantir side and then I can talk about our journey. I've reflected back a lot on that experience because it was deeply formative and there was a lot of learning that was happening, but it was also one of these companies that had a very unique perspective on how to run a company. And at the time, there were a lot of things that were like great about it and there were a lot of things that were frustrating about it and now several years removed, I think one of the things that in retrospect stood out to me is I showed up at Palantir, bright-eyed, bushy-tailed, straight outta Stanford, and they were like, we've never built web technology before. We've also never built for this particular customer or, you know, space before. You have half a designer, figure something out. I literally like one week previously had like, you know, walked off stage and gotten my diploma. Graduated.

It was like objectively a bad idea. Yeah, it's a terrible idea.

And it went predictably poorly. So it did not work out. We built something that was not adopted, but Palantir was full of those things. And a lot of companies I think would look at that and be like, that's a bad idea, don't do that. But at Palantir, there were enough of these things that some of them actually were great. And the sort of secret sauce I think that they had there was one, being able to find incredibly talented, hardworking, driven people, and then two, giving them leeway to go use those skills and giving them opportunities to apply those skills in ways that were very relevant for starting companies because there wasn't a lot of handholding. There wasn't a lot of top-down, like, this is what you have to be working on. They let smart people find interesting problems and then go solve them. And for us, where that comes into kind of our story is several years later when Foundry, the Palantir sort of platform, was in its infancy. On one end, I was trying to figure out like, what do we actually need to build for this platform? And on the other end, Barry was struggling with this problem with the customer he had. And we actually got to test drive being co-founders. I had this alpha that I wanted to deploy to a customer and we like flew up to Alaska and like sat next to this petroleum engineer and watched him use the platform.

Um, shout out to Brian.

Yeah.

Brian's listening to this. I think about you sometimes.

We still talk about Brian sometimes. He's actually like a, in some ways, a prototypical Hex user because he's very technical and doing sort of real data analysis work, but really was just like extremely poorly served by the tools of the time.

Most co-founders co-founder date. They either meet in a coffee shop or events like this. Sometimes they're teamed up, but some of the strongest relationships are ones where you have prior art around working together. Tell us a little bit more about the aha moment where you kind of test drove co-founding, but then you decided, okay, we're actually going to quit our jobs and do this full-time.

It's funny because we were fortunate enough to test drive co-founding when we were working without realizing it. It's not like we were like trying to test drive co-founding, we were doing our jobs. And then we actually, all three of us co-founders went our separate ways. Actually, Glenn and I went to the same company. I went first and then he followed me and then Caitlin went to a different company, but we all left Palantir for like a year and a half and did other things and then came back together. And so it felt very natural because we had worked together. Glenn was actually my intern, like in 2014. So we've been working together for a super long time. And I've observed this a lot through angel investing and talking to a lot of founders and seeing a good number of founder breakups now where it doesn't work or it's just like, it's clearly not a healthy relationship. I think one of the things that really helped the three of us is we were all, this sounds funny, we were like raised the same way at Palantir. We all joined in like our early 20s, we're like programmed and had stuck around and had some success there because I think that culture worked and it worked for us. And so we had this almost like values alignment. It wasn't just like, We think the same idea is cool or whatever. It's like we had a similar pattern that we'd been programmed with of like how to build software and how to run teams. And so just, I think if you like culturally, there's a lot of stuff that was easy that I wish more co-founders that were co-founder dating, like got into the really hard stuff early and like almost found ways to put themselves in joint stress positions, like to find out like, do you really have shared values on this? And actually when we hire leaders, like we just hired a head of people, I spent a lot of time with this candidate trying to find all the craziest, most effed up situations to talk about. Like, like, cuz you wanna feel like if you're really partnering with someone very deeply, you're gonna be able to navigate the hard times together. When things are good, everything's good. Those are the easy times. But like running a company kind of sucks. There's like a lot of ups and downs and I think having that grounding and like you kind of look at the world the same way on some stuff is really helpful.

When you pivot, which many companies do, it's important to have something deeper than just the idea that holds the team together. For the founders that are in the room, maybe looking for co-founders or just starting their companies, what advice would you both have for folks that are looking to build the founding team?

I have a, a whole thing about this that you talk to a lot of founders or prospective founders and they'll tell you their startup idea. And people jump to that really quickly cuz that's the archetypical thing you're supposed to produce. But the best companies are founded around like a startup problem. Uh, give this advice to someone a few months ago, I was looking at his pitch deck to raise a seed and it was like this really detailed, very like measy pitch. Here's the idea and here's the market opportunity and here's all the ways in which it's going to work and the existing things and like, here's how we're going to win and a 5-year plan. I was like, dude, no one reading this is going to believe all of that. And like, you don't believe it because you don't know because you're early and your thing sucks and like, it's just getting off the ground and that's cool. I think the thing that you should do is like focus on the startup problem. I gave him this feedback. I was like, redo your whole pitch deck, just selling the problem. Be like, I understand this problem in and out. It's a huge problem. No one's solved it yet. I totally get it. And what's interesting is I think that is actually what winds up attracting early investment more than your perfectly formed idea. Because even if you don't realize it, the VC on the other side, they know that things change a lot. They're looking for your understanding of the problem space, but it's also really attractive to candidates either as co-founding candidates or your early hires. People are attracted to problems. And I think if you try to explain your whole idea and your master plan, maybe they'll buy it, maybe they won't, maybe they'll poke all the holes in it that it deserves to have poked in it because it's an early idea and you don't know what's gonna happen. But if you just sell people on the problem and like there's an injustice in the world that like someone's gotta fix this damn thing, you're putting up the like beacon in a much more like vibrant way. And I wish more people did that. I wanna see more people do that.

That's a really helpful centering. And with that, I'd love to dig a little bit into the problems that what Hex is solving for. Caitlin, for folks in the room that don't think day and night about data and dashboards, though people probably should these days, what is Hex's core product and what are you solving for your core user?

Hex is a product for data scientists, data analysts, people who are working with data day in, day out. Increasingly, that ends up being people who also have data in their jobs, but maybe data in their job titles. And I think we can talk more about what AI means for that. When we started, Hex was really focusing on these folks. And the thing that I realized when I was thinking about signing on with Barry and Glen was that at the time, no one had really built a tool for these people that really connected their whole workflow together. If you looked at what these people were doing, they had Jupyter Notebooks over here and they had Looker over here, and it was all duct taped together. And then at the end of the day, they were still just screenshotting charts and copying and pasting them into PowerPoint slides. And that was just such a brittle, painful workflow that even just starting with this problem of like, this should be better, there should be at the time a modern cloud-based collaborative SaaS offering for these things, and there just wasn't. Focusing on that, and really, I think one thing we've done extremely well as a company over the years is staying laser-focused on that group and making sure that they are the number one Hex proponents and that they are loving our product every single day.

That's incredible. Hex started before this wave of generative AI, and we are at Tech Week in 2025, so we have to discuss AI. Tell us, would you consider Hex an AI-native company? How does AI infuse itself in your core product and strategy?

I struggle with AI-native a little bit because we were founded before this era of generative AI, but we have done a lot of work to make sure that we are AI-first and that we are really deeply engaging with the technology and again, solving the problem, the founding mission of Hex, what does it mean to be the best tool to work with data in 2025? And that we're really confronting that very head-on and we're not just bolting on some chatbot somewhere. One thing that we did earlier this year in service was that if you rewind 12 months, we had like a little AI team. They built our SQL Gen, they built a couple different features, and they were happily working along on this roadmap of AI stuff. And we took a look at that and we realized that at the speed and the pace that AI was moving and the new capabilities of the models at that time, that just wasn't sufficient. We actually had already published a blog post being like, AI is the product. We're never going to charge more for AI because the product is AI. And we had this moment of like, wait, we're not even really following that. We just have this AI team in this little silo over here. So we basically completely reorganized the entire engineering team to instead of have one team focused on AI and being the only people that could like touch the prompts to basically make it the responsibility of every single product team to not just own their own features, but also own the sort of product story around what does this mean in the context of AI and delivering that.

When you say it that way, it makes me feel like having an internet team today. Imagine saying that. I have an internet team.

Some people have never even heard of this product, which is funny to me. There's a product called Sketch that was like the design product in the mid-2010s. If you were doing product design in a tech company, you were using Sketch and it was this beautiful, really rich, really full-featured Mac app. And Figma came along and Figma was like a toy. It was like this cloud-based thing, but it didn't have so many of the features Sketch had. We could never use Figma, but like clearly like there's something to this cloud thing. And so Sketch started a cloud team. Figma never had a cloud team. It just like is, it's just like the water you swim in. And I use that analogy a lot internally. I was at all hands, like, how many people have heard of Sketch? And like a lot of people are too young to even remember it. And I was like, exactly. You've never heard of it. That's how they They're really smart people and they build a really great product, but they just got that thing wrong. And so I, we think a lot about that and a lot of the, the companies that are doing really well with AI now were founded well before ChatGPT, like a lot of companies you think of, Vercel, Clay, I could argue even like Cursor. It's like, it's attached to VS Code, which has been on for a long time. It's not necessarily like you had to be founded post-GPT. It's like you actually have to have this like deep intimacy with going back to what I was saying, the problems and the workflows and the users and like know how to sell to them and be able to bring AI into those use cases. It's not hard to build a vibe coding tool. It's very hard to build it in the way Vercel has with v0. It's integrated well into the rest of the workflows and to know what professional developers need. And that's fundamentally what we're doing with AI at Hex. The initial problem we were solving was very focused on like, there's this group of people and their workflow sucks. They want to get answers from data and it's really hard to do that. It's really hard to do it together. And in a way, nothing's changed. For Vercel, it's too hard to build and host websites. And like, in a way, post-GPT, nothing's changed. That problem's the same. The mission's the same. And for us, the kind of thing, fundamental thing we're trying to build is very similar in spirit, even though there's like this magic new technology that shows up. I wrote a blog post about how we got rid of our AI product team a few weeks ago. I was almost embarrassed writing it because I was like, this feels so obvious. But then I heard from so many founders being like, how did you make this work? Thank you for publishing this. This was brave. And I was like, I think a lot of people are probably still on the wrong side of this. And I, this is my call to action for people to get rid of their AI teams.

You mentioned the word magic a few seconds ago. Let's talk about HexMagic a little bit. What is HexMagic? And with building in the AI environment, tell us a little bit about the implications, whether benefits or issues with building an AI feature set so early on.

Hex Magic is what we branded our AI features when we first launched them, and we didn't wanna call it Hex AI. We're like, magic's tight. Like it plays into the Hex theme. It was like, like, I don't know, I like branding things. It was like, it was a really part of that name. We had a magic team and we had a little magic icon and we had like cool wizard swag. It was like really sick. But like, this is kind of my point is actually we killed magic. It's not ask magic to do this, it's ask Hex. Like this, the product does now. And I think that's like a mentality and it was almost as really important internally as externally to like send that signal. Now the core of your question is like, what's AI mean for data workflows? And I think the core promise of data tools and anyone here has ever like, or anyone listening to this that's ever interacted with like a BI tool or a dashboard, they're, they've been sold the same way for like 20 years, which is like, you'll buy this thing. and this will let all the people in your organization access data. And then in reality, that doesn't super happen. Some dashboards get built and shared, but like really, is everyone data-driven? Like most things wind up just being tossed over the fence to the data team. They're using a melange of tools. It takes them like weeks to get back an answer, build you a new dashboard, and then when you get it, you really just have more questions than you have answers. It's like, oh cool, our signups are trending up. Why? What's that correlate with? What segments are driving that? Maybe you have the drill-down ability, probably not. 90% of people are not going to be proficient enough with that BI tool and know where all the buttons and knobs are to like get further. And what you have with AI is this blindingly obvious thing. If I had gone back to when ChatGPT first launched and I was like, name 10 use cases you could think of for LLMs, most people are going to be like, oh, can this thing write SQL? Oh snap. Okay. Like we should build a data tool. And a lot of startups got started that way.

Why?

Because it's so blindingly obvious that like I should just be able to ask a natural language question and get a data answer back. And it turns out that is really easy to build a basic demo of and really hard to get right in the real world. And fundamentally what we are doing and building and oriented the entire company around is like, how do you actually enable everyone to have that ready, easy access to answers, be able to do it in a way that's trusted. Connects all the different people in the org because you're still going to have these data people who want to dig in, want to get into the code, understand what's going on behind the scenes. You're going to have the rest of the organization that just wants to be able to ask questions with data. You want all of them to be able to be used the same context and know that the answers are accurate and grounded in truth. And building that system is what we're doing. And we're just very fortunate that we kind of walked into the AI era with all the pieces. And as the models have gotten better, our ability to wield them and build that type of product is, especially just in the last few months, has been amazing.

I've heard from some investors recently when they've invested in a company that sounds like there's probably 5 other of those companies or 10 others. I'm like, what's the differentiator?

Well, it works.

It actually works. It's not just a demo and it's more important than, uh, putting.ai at the end of your domain name.

We do own hex.ai though. That's important.

Keep that.

Keep that. Good thing to have under your hat. Well, I mean, something is definitely working. You guys raised your Series C this this past May. Huge congratulations. Incredible investors on the cap table ranging from Sequoia to Redpoint, Snowflake, and of course Andreessen. Who was the first person to say yes to investing in you both?

There's a few answers to that question. One of my coworkers from Formation Bio, Jeff Silverstein, was like a scout investor. We toyed with him when we were thinking about staying in New York and doing the company. We thought about having him as our third co-founder. He's like the secret— I think there was like a Beatle who like played with them for a while in Liverpool. Like, he's like that guy, you know, like, so he gave us our first check and then Nimi at Box Group, I met her when I was in New York. She was the first one to really step up with a larger check. And then Amplify led our seed and we actually met Sarah through you. So it was Sarah Calanzaro.

So the person who will not let anyone forget that they sourced the deal at Amplify is actually Mike Dauber. I had actually been connected to Mike when I was leaving Palantir and I was like looking for my next thing and we had a mutual connection and he's like, oh, just to talk to Mike see if there's anything. And there wasn't anything that I was super interested in. And then as we were starting the company and we were talking about how are we going to get funding and who we're going to talk to. And actually, even before we were really thinking about funding, I was like, I know one VC. And I emailed him. I'm like, hey, I'm thinking about starting a company, so we should get coffee. Then we got connected to Sarah, who was a data practitioner, was very deep in this space. And she was the one who really did a lot of sort of the diligence around what is the problem that we're solving? Are we the right people to solve it? Do we have the right mindset and the right sort of way of, of going about this. And then ultimately, the person who ended up leading the round was Sunil at Amplify, and he's been fantastic to work with.

That's incredible. I think sometimes it does take that kind of like reach out to that person you kind of know, and then they put you in the room and then the door opens. And of course that was what got the ball rolling for you guys. And then ultimately at some point, these crazy VCs, they ask you to start a board and create a board, and then you kind of report to that board. And oftentimes we talk about the board and this third-party concept, but no one's really willing to peel back the layers there. But we're curious with you guys, how do you work with your board? What's that dynamic like?

Yeah, it's interesting. It's one of the many things you wind up doing as a founder CEO that there's no instruction manual for, and you realize everyone does it differently. And so it's very funny. The formal function of the board is like governance. Technically, their job is hiring and firing the CEO, but you can use them however you want.

Caitlin's like, yes, I know that. Yeah, I'm aware as a CTO.

Yeah, my job is hiring and firing everyone else. It's all rolls downhill. The LPs hire and fire them. That's all. Yeah, Uh, one thing I've come to appreciate is how much you have a board that's built of different pieces. Like you go to different people for different things. And like I go to Sunil with stuff that is very different than what I go to Sarah Wang, who's our board member from A16Z, or Linda Tong, who is the CEO of Webflow, who's our independent board member who joined earlier this year. I go to these people with completely different things. There's like a Venn diagram overlap in the middle, but like, that's something I have appreciated a lot is like over time you can add different people with different skill sets. And people stress out sometimes at seed or any series of their fundraising. Like they're making this big choice. I'm trading off between this person has this skill set and this one. I'm like, you can always go back. In fact, every round, basically more or less, we had funds pass on us that then one of them led the subsequent round. It's a longer game than just you have to make this spot decision. I think people get really stressed out on it when they're like raising their seed. But as you get further in, you realize like you're building a team. Of people and you can build it with like different skill sets. It's almost like you have little video game characters with like different abilities, like agility, armor, except that it's like the things VCs are good or bad at. Uh, yeah.

That's helpful to think about the board as an extension of your team. I love that framing. I think a VC would hope that's true, right? You do have to build that trust with your board as a founder. I'm curious, there's been a lot of topics in the news with things like the Windsurf acquisition where this existential question has come into play of like who does the board serve? Is it the capital or is it the employees? As a founder, how are you guys thinking about navigating that open question right now in the market?

Most of the time you're working with your investors, it's like you're all on the same team and you're like going for it together. And then there are these moments when you're fundraising, as an example, you're really excited to partner with this person, but like they want a low price, you want a high price. There's an adversarial nature because you do fundamentally like have different incentives. Personally, I don't envy the decision that some of these founders have had to make with these like weird-shaped M&A deals. I think it's all a bit of a weird short-term distortion. I do worry that like we've broken some type of social compact, but I also think it's a question sometimes in an M&A deal, like what are people actually buying? And if they're fundamentally looking for the talent and the core of the product that the company built or the business they built isn't being interested in acquired, like that is where that, like, I think incentive divergence starts to happen. And I hope we don't see a lot more of that. It is something I've talked with our board over time. We acquired a company earlier this year, actually. So like, funny enough, I've also been on the other side of it, like, and that conversation was very interesting because like, that's a different dynamic. The founder who came over, we were really excited to work together. We were super motivated. The day it closes, you want to be like super on the same team, but it was actually like almost like this four-part negotiation because he has his investors. I have our investors, it's me and him. You're playing telegraph. There's an army of lawyers in the middle. It is just very complicated because you do have these competing incentives. But on the other side of that negotiation, you want to be like locked in together. And so it's just a weird dynamic sometimes. And the only thing I've learned on this, and this goes probably for negotiations generally where you want like a high trust outcome, is just to be super transparent. No guile. Lay your cards on the table and you might like erode your negotiating position a little bit, but like you wind up with this really high trust relationship. And the company we acquired and the founder, we have a really high trust relationship. Partly cuz we like worked through that together in a really trusting way. And I try to do that with our board too. Like when we have tough topics, I'm like, I'm just gonna put this on the table and talk about it. And it's a little awkward, but you have your incentives. I have my incentives. We all know that we're all adults. Here's the deal. You can just do that and save a lot of stress, I think.

No, that's incredible.

I think when it comes to building relationships for the long term, there's a lot of, especially in the acquisition front, like, oh, great companies are Bought, not sold, and there's all this posturing and for founders that are building in this environment, maybe transparency is actually the thing that gets the deal done.

I think so. And like, I, I've seen a lot of this recently. Maybe you've hit the point where like you don't think it makes sense to continue as an independent company. In our market, there's like two really big independent players, there's Snowflake and Databricks. I talked to a founder recently who was like in our space. He was actually interested in us potentially acquiring them, but was also talking to other players. And I was like stepping out of the Hex thing for a moment, how you're thinking about that. And he was like, well, if I can just get these two guys to play off of each other. And I was like, they're very aware of that game now. That's, they know a lot more about being played off each other than you do. And my suggestion to him, and I don't know if this is ultimately good advice or not, it could be bad advice. I told him to take it with a grain of salt, was like, I would just go to them and put your cards on the table. And I think they're going to respect that a lot more than be like, well, you know, I've got a lot of interest from another player in the space. They're going to be like, like, screw you, I know what you're trying to do. I think if you go to them and you're like, look, it'd make a lot of sense for us to join you. I could sit here and try to pretend, try to ping pong you guys around. Here's a number which I would get this done and we'd be excited about this. I actually wonder, no one take this as like definitive advice, but I knowing a lot of the people on the other side, I actually think that would go further in them being like, okay, this is a high integrity person we'd be interested in working with than the like maximum guile. I'm gonna be master poker player type thing, if that makes sense.

Yeah.

Let's shift gears a little bit and talk about how Hex is scaling. It's been a very crowded market. Analytics, fast-moving, hot topic. Everyone's talking about data. Everyone's talking about AI. Maybe Caitlin, can you talk a little bit about how you think about product differentiation and ultimately continuing to build a quote-unquote moat in this current environment?

Speaking of values alignment and things that we really agree on as founders, one thing that we commonly like to cite is that there are no moats in software. There is actually a blog post that one of our colleagues wrote a while ago that talks about what the real moat in software is, is the cumulative knowledge and understanding that you build in the space, and then the ability to quickly take action on that. That's never been more true in the era of AI. I can go down and I can like run you through a list, like I can I've seen the decks that we present to our customers that like run through the feature-by-feature comparison about what Hex has and what these other competitors have. But at the end of the day, all software can be built, and that's even more true now with AI than it was ever before. But at Hex, I feel that we've built such a deep empathy and understanding for these users, and also a really deep understanding of the technology and what it's capable of. And really our biggest advantage is our ability to take those two things and combine that and move very fast to deliver, again, the best product experience for folks.

Part of your mission and vision is making everyone a data person, but in reality, not everyone is a data person naturally yet because they're not using Hex.

Because they're not using Hex.

So true. So true.

But I'd love to peel back the onion a little bit around your go-to-market.

How are you thinking about blocking and tackling, mobilizing a sales team, different methods to really make this vision a reality?

I think there are markets that can grow really well purely product-led. There are other markets you couldn't imagine doing product-led. You need like super thick sales. Data is somewhere in between. I think for us it's like a segmentation thing. We have a ton of smaller startups, thousands of smaller startups that come onboard to Hex, hook us up to their data. We never talk to them. They're successful. And there's a lot of much larger companies and enterprises where we have long sales cycles with them sometimes longer than I'd like, because it's, they're evaluating you in these long things. And especially now with AI, there's like a lot of questions. And so I've been thinking about this a lot and like one question that we've been thinking about is like, does AI change the way you do a go-to-market at all? I think a lot of people, like the first thing they see is like there are a lot of companies growing really quickly through pure PLG and AI. There's also a lot of companies growing quickly that have great sales teams. And I actually wonder if for us it almost counterintuitively means that we want to be even more engaged with the customers and not less, because in our space, getting your context set up right so you can get like accurate answers from your data is actually like super important and best that's something that we've always had a pretty good motion around, like partnering with customers. But Hex was always so easy to use and get started with. And now we're doing a lot on this now, like just digging in with customers a lot more. Because if you can get them to the aha moment where like the product is really working and they ask a question, it gives them a great answer and you're starting that flywheel. I think there's like this amazing compounding thing. And one thing we've always had be really strong is our expansion rates and NDR has always been like super high because we built a product that spreads really well. And it's quite viral. You went back and said like, not everyone's a data person. To us, that's the problem. That almost feels like this intimidating thing. Like, oh, you'll hear people say like, I'm not a data person. It's like, why? You don't know the whole pandas syntax? That clearly doesn't matter anymore. What are the real barriers there? And I think you're still going to have people on data teams. I think you're still going to have people in data roles. That's a great thing. Those people too want just everyone else in the organization to get a lot of value from their data and not be like intimidated by it. And so That's our vision. And we think that the things we're working on now are gonna 10x. People say 10x is such a cheap thing to say, but like, I said that the other day, like we're gonna 10x this workflow. And someone was like, that's like an off-the-shelf kind of hyperbolic thing to say. But it's like, that's probably underestimating it. The ability to just type a natural language question and get back like a full analysis is like, I don't know how to like even quantify like how much better that is than what you would've been doing before.

I'd love to jump in on that, like future vision state and thinking through in 5 years, maybe let's just say to keep it simple, like Caitlin on the back end? What's that gonna look like? You have all this hook set up, the data is just like magic for the end user. Like, where do you see us going in terms of data integrations and this future state of AI-powered data?

5 years. That's such a funny timeframe. It's like as long as we've been doing—

No, I know, I know. I'm asking you to predict the future here. It's Hex, come on.

Yeah, I'm not sure that I can predict as far out as 5 years, especially when I'm not sure, you know, what everybody's got cooking. OpenAI and Anthropic and things like that. But I do think that there's a couple pieces that are really going to change pretty dramatically over that time period. And I think historically there's sort of two sides to this coin of making this data ecosystem work. There's the people who are consuming it, and Barry just talked about that and how much more accessible that's going to be. But it also will probably be more present and more ambient, and it's not going to be like, oh, you load up a dashboard anymore. You're going to get an email in your inbox and it's going to be like, hey, the background agents that were running overnight found this discrepancy. Anomaly detection has long been a thing that people have attempted to build, and now that's actually possible with these things. And so on that end, and it's also is going to be present in every surface that you're working in. It's not just going to be, oh, I have to specifically navigate to like hexapp.hex.tech to like get all my data. That will be present where I'm doing my work. And then on the other end, I do think we're going to see a lot of change in how data teams think about their job. I mean, there's nuance to doing data work. And so people who really are experts in that will— there will always be questions that they need to answer. But increasingly, we're internally leaning into our data team and leaning into context engineering as not just a thing that they do, but maybe even the primary thing that they do, because even today, like, you can have the smartest model in the world and it's still garbage in, garbage out. And that's just going to continue to be true for these, like, really complex enterprise data systems. And so thinking about what that workflow looks like and how those people build the system that can be leveraged for the entire rest of the organization to actually get accurate data answers is going to be really important.

That's a super valuable framework because I think there is this, like, existential dread question that sometimes we all face of, like, where do the humans fit in if everything is automated away? And so from your perspective, sitting within the CTO org and looking every day at both like the engineers building the core product and then of course helping your engineering and sales teams integrate your core product into these third teams. How do you think about that magic between automation and human in the loop?

One thing that one of our head of data once said is, my job is not necessarily to answer your question, it's to help you understand what the right question to ask is. And I really think that becomes a core heart of the data team. Obviously they're enabling everybody else, but their ultimate goal is not to like answer questions. Their ultimate goal is to power the business. And still today, that's just a very human-driven process. And they're the people that are infusing that. Like, there are a thousand different ways to model a system depending on the types of questions you want to get answered out of that. And yes, those modeling workflows are going to be accelerated. We have our own data modeling agent. A lot of that drudgery, I think, is going to go away. But at the end of the day, the person who is doing that work is actually putting their own sort of stamp and opinion on how they think the business operates and what is important to know about the business. Business.

So humans, we're here to stay.

I think so. I, it's possible that data role evolves. A lot of data practitioners, if you ask them about the biggest pain points, it's like, oh, I'm constantly getting these like quick questions from people in the business. I'm answering these like kind of shallow questions and it's like, it's bad for everyone. Like I'm taking kind of longer than they would like to get back to them, but it's interrupting my work and I'm not able to focus on deeper strategic stuff. And so you can, I think, increasingly take a lot of that off their plate. In our space, in analytics, it's referred to as like self-serve, like let people do those people have a lot more time to do deeper things, more strategic things. I think a fallacy a lot of people fall into is they assume that the number of people employed at companies is some supply-demand balance on like, in data, there's like 100 data questions asked every week and each data analyst can answer 10 questions. So we'll hire 10 data analysts. That's not really it. There's effectively infinite data questions that people have. And then like the CFO only lets you hire so many data people. And then like, that's the number of questions that get answered. And there's this kind of like infinite demand for insight. And it's similar with software engineers. If you're like, oh, Claude code exists, therefore you need less engineers. Like, I don't know, like we're hiring engineers voraciously. Everyone is hiring engineers voraciously. It turns out if you make it easier to build software, you just get more software. That's going to be true in our space too. And there's actually like a weird world where you actually might have more data people than less, or like people in data roles than less, because they're just so much more valuable. They can actually like get back to you quickly with a thoughtful answer. I don't know for sure that's exactly what'll happen, but I think there's like like a sort of like base fallacy people are having of this very like fixed mindset, very like scarcity mindset thing.

Yeah, like we're tackling more and we're tackling bigger problems.

That's what everyone says they want. Uh, yeah.

Yeah.

They need to use Hex. Um, we spoke a bit about the Hex founding journey, how we're thinking about data and AI 5, 10 years from now. I'd love to shift gears a little bit and chat about what's next for both of you personally as builders and leaders.

I've been spending a lot of my time really deep in the bowels of this Context Engine and really understanding the, like, mechanics of that context flywheel that we've been talking about. And there's a lot, there's just so much there to unpack. There's product, there's technology, there's like frontier research. We have a small but mighty AI engineering team of 4 people and they're doing some really cool stuff. And so really like pushing the boundaries of what is possible there, but also trying to see a few steps ahead and understand really where that workflow is going. Because again, we think that's going to be be a huge part of the engine that actually like makes these products successful.

Yeah. As the company grows, the CEO job kind of gets interesting in new ways. There's like new stuff I have to deal with. We're at like 150 people, so that's the classic like Dunbar's number. Trying to think more about like the systems we're building. I can't manage everything by hand or be as involved as many things, like being very deliberate about where I want to be really involved, where I want to delegate stuff out. I think I'm getting better at certain things. There's other things that I feel myself still being bad at. And every day we hire a new person that's the new largest company I've ever had to be CEO of. So that's cool. It's like an interesting, fun challenge, and I'm very fortunate to have a lot of other CEOs I can reach out to and lean on and often feel better because when I text them with weird questions or things I'm stressed about, you realize that everyone's stressed out about the same stuff. So you're not alone. You're not an idiot.

Yeah.

Well, we're so excited to stay apprised of the Hex journey and see how the company continues to grow from 150 to maybe 1,500 very soon. And as we come up on time today, we'd love to ask our hero question of the podcast. That question is, who is a woman that has had a profound impact on your career and your life? Caitlin, maybe we'll start with you.

The, like, more normal answer is Julie Zelensky, who is one of the professors at Stanford and who was my advisor. And her course was really hard, like really hard. And I ended up doing pretty well in that class. And I think both that combination of seeing that and like going through this thing really hard and coming out the other side and being like, oh, I'm like actually good at that. And having her there as a role model was instrumental. Instrumental in me having the confidence to really lean into something. Like I, I loved coding, but it was always this sort of like, oh, it's like this fun side thing. And I was like, oh, actually, like I can go out and do this.

Incredible.

I had a boss, one of the founders named Mel, Melody, and she, I look back and cringe cuz I was like young and like a lot of energy and like was like making a lot of mistakes. And she not only didn't fire me, she like very patiently pulled me aside. And there's one conversation I remember in particular that she took the time to like really try to understand me and help me understand the bigger picture and zoom out a little bit and particularly like understand and like own something that hadn't gone very well. And I'll never forget that. That was very helpful. And I'm actually gonna, I'm gonna see her next week.

Shout out to her.

I don't know if I've ever told her this. Yeah.

We should send her this episode.

I will send her this episode. Yeah.

Amazing. Thank you so much for joining us physically in the room today. It's been a pleasure to have you here, hear your story. Also a huge Thank you to Perkins Coie Mercury. We have an incredible happy hour happening after outside, so please grab a drink, grab a bite. We also have some snacks and goodies from past podcast guests like Unreal Snacks. That's a great episode to listen to. So chat, enjoy, and thank you so much for being here.

Thanks everyone. Thank you so much for joining us at The Room Podcast. If you want more from The Room every week, subscribe to our newsletter at theroompodcast.com/newsletter. We'll be back next week with a new episode, an inspirational guest, Tuesday, 10:00 AM Eastern, 7:00 AM Pacific. See you in the room. Perkins Coie supports the most innovative entrepreneurs and investors in fast-moving and high-growth sectors, addressing their myriad of legal needs. But the firm doesn't just provide end-to-end legal and business counseling to its startup clients. It also facilitates introductions to key advisors and sources of capital. Perkins Coie's interactive website, Startup Percolator, offers access to programs, resources, and rich dynamic content designed to assist entrepreneurs on their startup journey. To learn more, go to startuppercolator.com and perkinscoie.com.

This podcast is brought to you by Mercury. Mercury, the banking platform businesses like The Room Podcast use to simplify their finances. It's time banking did more than hold your money. Now it can. With Mercury, you can pay bills in seconds, close the books faster, and even send invoices. Not only does Mercury do away with a patchwork of tools, it eliminates guesswork, giving you complete and accurate visibility into your business's finances, all from one account. Apply in minutes at mercury.com.