The Exchanges

Q Okay, so we did this interview with Tom Mueller who was the first employee at SpaceX and we were talking about data centers in space and we put out the videos and in the YouTube comments, you know, there's a lot of fear-mongering about birds and killing birds with the lasers and all that kind of stuff. So can you dispel some of the misconceptions about lasers and their strength?

A And look, it's all about power. I'll give you a really good example, right? The iPhone with the face ID, That's a laser product that beams lasers onto your face. And, you know, one of the concerns of course Apple had was, hey, I want to have people's eyes burning out, right? So the power of those lasers are relatively modest and you're never going to burn somebody's eye out. Similarly, you're not going to have birds burning up as these lasers shoot data down to the, to the earth, right? It all has to do with the power. You can have these, Super high power lasers that can shoot down missiles that can cut through metal. You can have modestly powered lasers that power the internet backbone, and you can have these super low power lasers that do things like the face ID.

AI assessment note: “Similarly, you're not going to have birds burning up as these lasers shoot data”

Q all about spending smarter, moving faster, but I take this in a personal lens. So I'm really curious from your standpoint, you've built up a tremendous career. You've been CEO of three public companies. I like to think performance component of that is like who you surround yourself with, who you're mentored by, who you're inspired by those types of figures along the way. Who are those figures for you?

A Look, I've, I've been super lucky. I mean, um, you know, you, you talk about Brex and performance, and I think you, you've got the right theme and the right partner for what you're doing. Um, I've been, been very lucky. My father, interestingly enough, was an electrical engineer, right? He was one of the first, uh, Electrical engineers that came over from Europe. He's actually, uh, was based in the United Kingdom and came over just shortly before I was born. And he had tremendous influence. He told me as I was growing up, you can have any major you want in college, just as long as the last word is engineering. Electrical, mechanical, civil, I don't care. Gotta be an engineer. And he had tremendous impact, I think, on me. And then as I got in my career, I spent a lot of years at a company called Broadcom, which I know you've covered pretty extensively. Broadcom was a very interesting company in that it was bought by Avago, and then the name stayed, the Broadcom name was the surviving name, but there was a gentleman at Broadcom, a guy called Bob Rango that I worked for for years and years, and Bob served on several boards and had a long and prestigious career, but he had a lot of Impact on me sort of from as a business guy and how I think about, uh, business problems.

AI assessment note: “there was a gentleman at Broadcom, a guy called Bob Rango that I worked for”

Q note, and so one of our long-term Partners is Brex, and they're all about performance, spending smarter, moving faster, and I believe a large part of performance is who you surround yourself with. I mean, you have quite the career, and you've been surrounded by industry legends, but I'm really curious from your standpoint, who are some of those people that have mentored you or inspired you along the way?

A There are quite a few, right, ah, I go to different mentors or leaders for different type of advice. When, because I started my career at Oracle, ah, what I saw with Larry Ellison and the team that he had at the time, ah, very focused on always creating the best product. And also very focused on having the best go to market teams as in How do you sell? How do you serve customers? Top notch. Like they would have this bar on talent or you should only hire software engineers from these colleges in the United States or in the UK or wherever the case might be. That's what they would do and that's how they would operate. So you learn quite a few things on how they would do communication in the company and so on. So that was my first learning. I also learned a lot when I was at Symantec. Great CEO. Uh, his name was John Thompson, uh, and legendary CEO. Uh, he had a very storied career at IBM and then became a first time CEO at Symantec. And how do you work with customers? How much do you, uh, listen versus how much do you talk? Because everybody wants to pitch their products and he definitely had a way with customers that I learned. He also dressed really, really well. And so, uh, because he's like, hey, if I'm showing up in front of customers, I want them to know I respect them, and I'm going to show up with my best self, right? So that was John Thompson. And then ServiceNow founder,…

AI assessment note: “There are quite a few, right, ah, I go to different mentors”

Q For people who have not seen your lectures just yet, highly recommend it. Who have you spoken with? I know you brought on different guests and you bring them for the students. What was the plan there?

A Oh, you know, we had a, we had a great lineup. It's a, it's a labor of love. And, um, I was really excited to bring back some of those conversations back to school. Um, you know, if you're a student and you're making a big life decision, I wanted to make sure that you saw the whole AI stack from chips to data centers, to models, Um, and the, ultimately the AI applications. Uh, and so we had a great set of speakers across all of the parts of the stack. We had folks from, uh, NVIDIA and Grok. Um, we had Ali Gozi from Databricks. We had Tuhan from BaseStand. We had Sachin from OpenAI, the folks from Anthropic. Uh, so it was a great lineup and we're very lucky to have them.

AI assessment note: “We had folks from, uh, NVIDIA and Grok. Um, we had Ali Gozi”

Q Before you've had a customer come over to you, I would assume, what has been the worst travel story you've ever heard?

A It's actually my travel story. This is the travel story that started Navan. I was one of the first ones to start an offshore operation in Ukraine, in Odessa, Ukraine, and I went there, arrived there in the middle of the night, freezing cold. By the way, I hate the cold. Freezing cold. Came to the hotel because of some credit card issue, ah, the hotel canceled my reservation, and they told me to go down the street and find a new hotel. The travel agency didn't pick up, the software didn't work. I walked with my luggage, like, down the street, one, one hotel after the other, until I found something, but this was by far my worst travel experience, and really what inspired me to start the company.

AI assessment note: “It's actually my travel story. This is the travel story that started Navan.”

Q So both Broadcom and SpaceX are having a bit of a moment right now. I guess if we look back at Broadcom to your point, you helped scale that company up and it reached a massive inflection point. So what were the biggest lessons that you learned there?

A Yeah. The biggest lessons is each size you are as you grow requires a different behavior from the company. When you're 10 people, It's a network of us just talking to each other. I need very little processor systems. It's smart people just doing the right thing. You get to a hundred and you need a little more process, but not too much. And then you get to a thousand and you need more process. But again, not too much because you can like constrain a company by putting too much process in too early and stunt the growth. So you always have to assume smart people, stupid systems, let the right people do the right thing in the moment. But track them, how they're doing it, what they're doing, and make sure they're accountable for the actions they're taking. And then you get the 5010 thousand. It's like everyone knows the piece they know, but they don't understand how a whole business works. But bringing that information together and presenting it to the top level people so they can understand how the whole company's operating together is kind of the focus. And that was both Broadcom's growth path. And when I joined SpaceX, it was again about a thousand employees producing about one booster a year. And through the years of growing up Warp Drive, the custom platform we built there, we scaled it up to 40 boosters a year and 9000 employees across multiple locations, and it was just that …

AI assessment note: “The biggest lessons is each size you are as you grow requires a different behavior”

Q significantly changed with AI and automation, the way you scale businesses now is dramatically different than before, throwing more headcount at it, but now we're Throwing really talented head count at it, and then also automating some of that back end with, you know, technology, AI, all this kind of stuff that's happening. So how do you imagine companies now scaling up with this new mix versus just purely people?

A Yeah, yeah. Something like SpaceX and Broadcom, AI didn't exist, and so it just took longer, and it was harder to build the systems and the documentation. And figuring out where the automation opportunities were, and to digest the data to figure out where the trends were. Like, uh, at SpaceX, I had grown up to having a team the size of about a 175, building the core platforms. I've got a team of six here. I don't really think it's going to scale much more than maybe one or two more people, and I think I'll be able to do exactly what I did at SpaceX with only six, because AI, It drives most of what we're doing, and AI is now starting to enable things we weren't even capable of doing while at SpaceX. So that's one of the areas, and then you just look at machine automation or robotics and the ability to cost-effectively apply it. Like, AI is just driving the cost down and the speed up associated with finding unique ways to apply technology to help humans do what they're great at and getting the mundane, boring tasks out of the work that we do.

AI assessment note: “I'll be able to do exactly what I did at SpaceX with only six”

Q Co-design has become an important part of how do you create that kind of platform for the next three years because you're, you're building for the next three years. So how do you think about that and how do you go about chip design?

A Well, I, I think historically you, you made chips and you ran software on them and, uh, There wasn't surprisingly a close interaction because there was a layer called an operating system that lived between the chip and the software. And so, uh, you know, Intel and AMD made chips and people wrote software for the operating system or for the chip. AI has gotten so large and, uh, speed is so important that, that what they're doing is they're thinking about sort of the design together. That what changes could we make in software that would advantage the hardware or as we're designing the hardware, what changes could we make that would make the software easier to run? And so they're being designed sort of at the same time. And like anything, when you sort of design things together, um, the, the advantages are enormous. And, and so this is something that's really taken shape right now. Um, you know, one of the advantages of our relationship with open AI is we, we get to see exactly where the frontier is going and we get a chance to, to roll that into our designs. One of the advantages Google has is that their TPU can be designed in collaboration with the team building Gemini or the team of DeepMind, and so they can, they can inform their choices back and forth, and that's an enormously powerful thing that is surprisingly relatively new in our space.

AI assessment note: “they can inform their choices back and forth, and that's an enormously powerful thing”

Q Discovering amazing things. So in terms of investments, how has that led you? I mean, you've invested into SpaceX, Uber, Anderle, Flock Safety. I can name like four, 40 companies maybe right now, but like you've invested in so many companies. How is that and what are some stories?

A I think I like to talk about Niantic Pokemon Go as one of the whys. So the first thing I noticed was that these people were playing this game called Ingress. My friends were playing it. I started noticing other people playing it, but it was a fringe game that not a lot of people were playing. But the behaviors that they were displaying, they were renting helicopters, they were chartering boats, they were ditching their friends at DEF CON in Las Vegas to go out into the desert to claim these invisible things that didn't exist other than in this app. So I started watching it, and I was like, that is remarkable. What could get people to do this? And so I started playing it, and I started learning about the mechanics, and, um, I immersed myself in it, and then I started asking myself, why is Google doing this? So I realized, oh, Google ultimately is a mapping company, They're gathering points of information. They're getting these players to take pictures of it, categorize it, tell them exactly where it is, and to scan that, that, that information. It's basically like free mapping data. And once I realized that, I was like, this is really clever. I wish I could invest in it, but it's Google, so I can't. And the minute that alphabet spun out, I got all of these messages from my friends and they said, you realize they're spinning out Niantic and it's going to become its own entity. An…

AI assessment note: “I think I like to talk about Niantic Pokemon Go as one of the whys.”

Q Turing. Turing builds realistic reinforcement learning environments and data systems based on real operational traces, the kind of infrastructure Frontier Labs need to train superintelligence. Visit Turing.com slash S-O-U-R-C-E-R-Y. Talking about the public markets now, you have a public market company, Nebius. Let's talk through that because that is like a really hot company right now. They just had a huge deal. What was it like? Twenty six billion dollars.

A Yeah. We announced a big deal with Meta. Yeah. First and foremost, as a venture investor, I wouldn't wish it on anyone to own a public stock. I mean, having a stock apps be like the major app that I use on my phone now. It's very stressful having to mark to market every day. Uh, You know, that's an, an amazing story of an entrepreneur, Arkady, who actually built Yandex, which was built to about a thirty billion dollar market cap on the NASDAQ. And then the war broke out in Russia and many of the engineers followed Arkady out of Russia. And he basically collected where he was and was thinking about what he was going to do next with all these engineers that followed him. And it was actually one of the founders of Excel, Jim Schwartz, who introduced me to Arkady. And basically said, he's trying to think through what's next. He's got a few balls in the air. You should talk to him. And so for a couple of years, Arkady and I went back and forth on what would eventually become Nebius. And actually the early innings of that was me trying to convince Arkady to sell me his 30% stake of ClickHouse, which he rightfully refused to do. And, uh, but as I got into that negotiation with him, I got to know him as an individual and he is the most Quietly humble killer I've ever met. He is truly, truly special and end of one entrepreneur. And, you know, most people would have hung the cleats up an…

AI assessment note: “We announced a big deal with Meta. Yeah. First and foremost, as a venture investor”

Q Nebius is, could you just give like a brief quick line on that? And then also where does it fit for inference people? This will be a wide-ranging audience, but in terms of inference, why is that becoming so important now? Obviously we're seeing it with agents and more different types of compute that they need and all the data that they're creating, but like, could you just share that?

A Yeah, absolutely. So, uh, at its most base level, Nebius is intending to build a next generation hyperscaler. So they want to own everything from the data center itself through to designing the server racks. They actually have a hardware team inside the company. To building all of the software that goes on top of that infrastructure. So we talk a lot about how many data centers need to be built, how many GPUs need to be delivered to handle the demands of AI, both training as well as inference, which we'll get into in a second. These guys are one of the most aggressive teams to scale that infrastructure and meet the moment, and they want to own all of it vertically integrated, and that's really important when you think about delivering inference. So inference is we've now trained all these models and we need to get value out of them. So in its most basic form, if you're asking a question of ChatGPT and you're getting an answer, that's inference against ChatGPT, against OpenAI's model. But as you allude to, agents start using inference in an exponential way. The demand curve goes way up for what we, what we're going to need from an inference perspective. And when you control capacity by owning the data centers and the GPUs, and you control the software, you're in an advantageous position to be able to Uh, meet the moment and, and deliver inference for a market that, you know, we …

AI assessment note: “Nebius is intending to build a next generation hyperscaler”

Q No, they're not, Dylan. It's literally the config, like, Figma conference. Okay, so how many years have you been doing this, and why start it in general?

A Yeah, so we started in 2020, uh, right before the pandemic. And it was, this is actually our 10th config, because we've done international configs that are fast follows to the config in SF. But the reason it started, um, we had basically user groups starting to pop up all around the world. And then from there, as the user group started to really gain momentum, we wanted some way to bring everyone together. And, uh, at first we're like kicking around ideas like, oh yeah, maybe like a few hundred people. We decided, let's go big, like, 800 people, and that sold out instantly, so then from there we were planning to, right afterwards, like, okay, next time it's gonna be really big. Well, it actually ended up being even bigger because of the pandemic, and so it was like, this is gonna be a remote conference now.

AI assessment note: “we started in 2020... the reason it started... wanted some way to bring everyone together.”

Q Are there any speakers this year you're excited about?

A So, so many speakers, um, that I'm excited about, but two that I'm, like, off the wall excited about, um, one is Holly Herndon. She's a total visionary artist, too. I think people don't appreciate, like, how far ahead she sees things. Um, you know, she was Into very specific, like, quirky ways of using NFTs, or they weren't called it then, uh, early on. She open sourced her voice with AI early on, you know, this is back in, like, twenty-twenty-one or something, twenty-twenty. I mean, it's just, like, the, the time scale that she's operating on is different than the rest of us. Yeah. Um, Grant Sanderson, three blue and brown on YouTube, is another person that I think will give a great talk, uh, and I can't wait to watch it. He's, Someone that I've been following for a long time, I'm lucky to call a friend, and he does these explainer videos for math that are just incredible. And, um, uh, really uses graphics and design to be really craft forward in his explanations, and he puts so much care into them, and just has been a teacher for so many.

AI assessment note: “two that I'm, like, off the wall excited about, um, one is Holly Herndon.”

Q very valuable is who you surround yourself with. You've been at so many legendary funds at this point. You've been surrounded by the top investors over and over again, but I'm really curious from your standpoint, like who has been kind of like a mentor to you? Like, who do you keep, like, As a guiding figure for you as you continue to, like, evolve and grow in your career?

A Yeah, I, I mean, I wake up every day and I, like, pinch myself because I, I've just been so unbelievably lucky and fortunate to, um, as you said, work with a lot of the people that have, like, defined the history of, of, you know, venture capital and growth investing, um, and not only have they been extremely good at their job, but a lot, like, basically every single one of them has been a wonderful person. I feel like I, I learned something different from every single one of them, and so, like, if you take, like, a Napoleon Ta Who is a GP at Founders Fund, who runs the growth investing practice at Founders Fund. Um, I mean, I learned a ton about investing with him, but the most important thing that I learned from Napoleon was his focus on family and work. And he had a very simple life. He doesn't, you know, he doesn't do, you know, he doesn't do podcasts. He doesn't, you know, he doesn't network. He doesn't, um, you know, try to spend a bunch of time with other investors. Um, he goes to work, he works really, really hard, and then he spends time with his family. And it's like this beautiful, simple life. And he loves, you know, both spending time with his family and working with the Founders Fund team. And it was something that, that made me realize like, wow, uh, and just like helped me prioritize just my time and how I think about the important things in my life. Um, I think…

AI assessment note: “like if you take, like, a Napoleon Ta Who is a GP at Founders Fund”

Q Who do you think your biggest competitor is?

A Um, it depends, uh, in what category, because Coinbase does a handful of things. I think, um, you know, if you look at, uh, stablecoins, like we've partnered with Circle on USDC, the biggest competitor to that would be Tether, right, which is offshore. I think, um, on the exchange, like on the crypto exchange side, our biggest competitor is probably Binance, right, which again is offshore. Um, I think if you look at the US market where some of these, Uh, fintech apps are integrating more crypto, you know, and we offer stock trading now. So Robinhood would probably be for our retail apps specifically, like Robinhood is probably the most, um, clear comp. And then, uh, with Coinbase developer platform, which is what it's allowing all these businesses to integrate stable coins, you know, Stripe is probably the best competitor we have there. So, um, yeah, we also run the largest institutional business in crypto. I actually don't think there's like a close competitor for that, at least not One that comes to mind. So yeah, we have a number of different products and, um, competitors in each of those categories. Oh, by the way, one other maybe interesting take on the AI, uh, usage, you know, you're asking about anthropic and all that.

AI assessment note: “our biggest competitor is probably Binance... Robinhood is probably the most, um, clear comp.”

Q the book, you literally started the book with Elon. He's very hype in the news cycle right now with the SpaceX IPO. He didn't need more press. He didn't need my help. He didn't need more press, but you started the book with him. Why do you think Elon is so emblematic at the speed of play? And then also, like, what was the process of putting that in first?

A Well, the process was It just made sense to me. It just, every time I came back to Like, what did, I wanted to get across the essence of life at the speed of play of this concept. Why did I put those words together? And why is that, why does that resonate so much for me? And I thought if I want people to picture it, I'd like people to picture the kinds of crazy things that Elon can go do that none of us can do and crazy from one hand, like putting fart mode in his car, which No responsible corporation would ever allow a founder to do, um, to serious things like, you know, deciding he wanted to create a tunnel boring company while he was stuck in traffic in LA, and we all have those thoughts, those moments. I don't know about fart mode, but some version of it, and None of us can actually do anything about it. Or if we do, we have to like devote our lives to that one thing. Like if you want to create a tunneling company, you know, maybe, maybe you or I could, but it wouldn't just be a tweet, right? We would have to like do our homework and go and research that industry and possibility and create a business plan and blah, blah, blah, blah, all the things that are not fun. And he's just having fun. You can tell watching him. So that's, That was the, the first part of your question, why I wanted to start with that. And then process wise, it actually wasn't easy because I don't live …

AI assessment note: “why I wanted to start with that. And then process wise, it actually wasn't easy”

Q missed the whole SaaS wave, but starting off in AI, it was all chat based. Now we've completely moved to the agentic era, which is like a thousand times more cost competitive. You need compute power, memory, All these sorts of things that are reorganizing how companies are being built and of course the stock market too. So how are you thinking through the agentic layer and building that out?

A Yeah, so I think this was like a huge transition we've gone through in like the past six months where exactly like you said, the original product we built was kind of this chat based co-pilot for lawyers and then we built all this product around it that Let it work at a law firm on client matters and all these things. And then maybe six months ago when these coding models started getting really good and you could do kind of things like cloud code and codex and that started working really well in the command line, we started building a bunch of the infrastructure of, you know, how do you get these agents that were running in your command line and could execute all these tools and work really well and how to move that into the cloud. And I think now you're seeing this with kind of these managed agent solutions. Um, and to your point, it's like now you need sandboxes. The models use way more tokens. Um, we have queries that like, we have simple like assistant type queries where you say, you know, draft me a document that a single query can cost 20 dollars. We have, um, like a review product where you can upload a 100,000 contracts and ask the models to review them, and some of those can cost 20,000 dollars. And so it is just getting Incredibly expensive, and I think in parallel what we're seeing is I think even two years ago the models were so good with coding that most programmer…

AI assessment note: “we started building a bunch of the infrastructure of, you know, how do you get these agents”

Q What are the top performing types of content that you have? Is there like a specific kind that like just always works?

A Uh, I have a series that I do called Hot Startup Rounds that has done really well. I think Contrary to, like, what we know as our daily jobs and, you know, it feels like we live and breathe this stuff, a lot of people actually don't feel like they have visibility into startup fundraisers and where money is going. Like, they'll hear about the open AIs and the anthropics of the world raising large rounds or, you know, pre-IPO stories, but actual visibility into the earliest stages of building, um, I realize that a lot of people are actually very interested in that, right? I, I think in what's being seeded, early stage ideas, and Oftentimes it's actually interesting. I'll get comments like, where do you hear about startup fundraising? I'm like, it's a no-brainer. TechCrunch, like textbook, Crunchbase. It's all stuff that feels very intuitive to us. But I started externalizing a lot of that, and so once a week I would share two or three startups that I think were notable. They didn't necessarily have to be, you know, the largest fundraisers. It could be interesting tech. It could be interesting stories. Those videos have done really well because I think people care about Genuinely, what's being funded? What type of technologies are emerging? Uh, and then I do a lot of content around breaking down general tech news. I would say, uh, usually with the angle of what is this? Why does i…

AI assessment note: “I have a series that I do called Hot Startup Rounds that has done really well.”

Q So are you going to keep doing Zowie talks in this new role? Like, how is this being set up?

A Yeah, uh, that was a really important point for me was that I think as even an investor generally, like, you should always have an opportunity to Hold your views, um, independently and have editorial control of, you know, what you say and, and how you think about the world, and I, I think most venture funds, and even at a place like Lightspeed, having differing views is welcome. Like, it's what makes us all better, and so I think for me it was really important having Zowie Talks as the space that I could continue to refine my ideas, continue to interface with the public sphere, uh, in the same way that I think when you have In the VC ecosystem, a lot of people have independent X accounts, um, or they have independent blogs. Like, those are places where they can share their independent thinking, and that in turn is actually, I think, enriches the overall firm's, um, collective understanding of, you know, Nuance and also understanding of different opportunities and, like, I think nurturing of, like, healthy discussion, and so it wasn't honestly even a point of discussion. They're like, oh, obviously you'll, you'll continue to have your own platform in the same way that I think if you look at a Fred Wilson at USV continues to have their own blog or the various exit counts of, um, investors in the ecosystem or, like, Bill Gurley and BG too, or sorry, um, Brad Gershner And formerly …

AI assessment note: “it wasn't honestly even a point of discussion. They're like, oh, obviously you'll”

Q And why did you go after the secondary angle? How did you have conviction in that early on?

A I think the insight that we had, if you kind of roll back the clock to like, 20 11, that was right before Facebook had gone public, and you know, our belief was companies were gonna stay private longer. Facebook stayed private for a relatively long time, but the major reason why Facebook ended up going public was this 500 shareholder count rule that basically required companies Once they hit a certain shareholder account to publicly report. So it was sort of all the negatives of being a public company with none of the positives. And ultimately that's why Facebook went public. They changed the rules with the jobs act. And so that pressure to become a public company sort of went away. And I think our view was very simply that companies were going to end up staying private longer. And what that meant was there'd be a lot more opportunities to invest in those businesses. Um, and especially if they were great companies, then That was just a lot more time they could compound. Going back to the SpaceX point, like SpaceX always sort of allowed liquidity for their employees and they were running tenders, you know, once a year probably back then. Um, and then that kind of sped up to twice a year. And so like these were always company organized programs. It wasn't like you're trying to find some random person who happens to, you know, be affiliated with the companies. Like you're just dea…

AI assessment note: “our belief was companies were gonna stay private longer”

Q Have you experienced any of the launches with any of the SpaceX team?

A Oh yeah, I mean, my friend was one of the engineers at SpaceX, right? I mean, I, Like, you know, there were some of the, you know, the old launches, like I remember going down to Hawthorne for like the third Falcon one launch. That was like the first launch after, um, we made the investment at founders fund. And so actually that flight, I remember flying down there cause you never quite know when the launch is going to happen. And like the flight was like slightly delayed and whatever. And I'm literally like running into the building literally as there, there's like, Like, they, I think they said one, and then like the rocket took off. Like, they were like, they were like 10 minutes delayed, which was perfect, um, and I caught the launch, and you know, that was the one where the second stage had hit the first stage, and so, you know, it wasn't fully successful, and I remember, I think I was standing next to Gwen or something, and she, you know, I was like, oh, you know, what, so how are you guys gonna fix this? And she was just like, I think we just need to add a delay between, between the, you know, on the stage separation, um, and like Steve Davis, I think, said it was like seven and a half seconds or whatever, um, And, and so, like, that was it, right? Like, that was the only change they had to make, like, and then on the fourth launch, everything worked, worked perfectly. S…

AI assessment note: “I remember going down to Hawthorne for like the third Falcon one launch.”

Q And why did you go after the secondary angle? How did you have conviction in that early on?

A I think the insight that we had, if you kind of roll back the clock to like, 20 11, that was right before Facebook had gone public, and you know, our belief was companies were gonna stay private longer. Facebook stayed private for a relatively long time, but the major reason why Facebook ended up going public was this 500 shareholder count rule that basically required companies Once they hit a certain shareholder account to publicly report. So it was sort of all the negatives of being a public company with none of the positives. And ultimately that's why Facebook went public. They changed the rules with the jobs act. And so that pressure to become a public company sort of went away. And I think our view was very simply that companies were going to end up staying private longer. And what that meant was there'd be a lot more opportunities to invest in those businesses. Um, and especially if they were great companies, then That was just a lot more time they could compound. Going back to the SpaceX point, like SpaceX always sort of allowed liquidity for their employees and they were running tenders, you know, once a year probably back then. Um, and then that kind of sped up to twice a year. And so like these were always company organized programs. It wasn't like you're trying to find some random person who happens to, you know, be affiliated with the companies. Like you're just dea…

AI assessment note: “our view was very simply that companies were going to end up staying private longer”

Q So what does scale actually mean for impulse?

A Yeah. Being able to simultaneously execute a number of things. So, you know, we are, uh, uh, we've got our Helios program, uh, which is the build effectively at the third stage for medium launch vehicles. That's building a, uh, rocket stage. You know, it's like, it's roughly the same size as like the second stage of Falcon one, for example, like as a size of the, the, the program, that's a big undertaking. We've got our, our Mira, uh, product line, which is now scaling to where we have to build one of those per month, uh, at this point. That's a big deal because we both need the engineers to develop it and the production to get it done. Um, and then we've got aspirations on building a number of other vehicles as well. Um, and so being able to simultaneously do multiple things is not easy for a company. Um, so scaling for us is really how do we hire great engineers across all of those programs? Um, there's a saying in, in, uh, in the Valley that, uh, you know, startups don't die from, uh, starvation. They die from indigestion. Right, and I think it's a balance that we always try to find the right place to strike of, hey, should we actually go out and try and execute on this amazing business opportunity, or should we lead by the wayside? And as I think we've maybe talked about a little before, there's not that many great opportunities in the space business segment, and so when we…

AI assessment note: “Being able to simultaneously execute a number of things.”

Q I've had over the last two, three years was, uh, you can see when companies don't really make a concerted effort to pivot or rebuild or, or embrace AI, and you can tell either by what they're posting about online, who they're engaging with, like how their products are developing, but you guys did make that concerted effort. What was that process for you internally, especially with your customers too?

A I mean, it was really hard. Um, there were a lot, there was specifically one really, really big deal that we wanted to close, and it was going to take up, like, 90% of our resources, um, and it was really hard to figure out, like, okay, how do we end up, like, taking, how do we end up finding resources to start building for the future instead of just building for right now, and thankfully, even though it was honestly at the time really hard, the deal paused for a month, and during that time, we actually got a lot done. We, we, like, finalized the idea that for our, like, Second product, which was the Merge Agent Handler. Um, we also started really aggressively leading into AI coding, so Gil and I, like, paused back on the keyboard, like, using Cloud Code really aggressively, using Windsurf really aggressively, so we could also learn how to code, um, with AI, and that, that month actually really transformed the business, because with us having first-hand knowledge of how powerful AI is from zero to one, it allowed us to see what was possible with our existing products as well.

AI assessment note: “during that time, we actually got a lot done. We, we, like, finalized the idea”

Q a lot to talk about. I know there's a lot of stories. I want to start with one of the stories that we talked about beforehand that caught my eye and I was Just flabbergasted by it. I just don't, I don't know how this happened. So you were on the Alibaba board. You were sitting there with Jack Ma, Joe Tsai, Masa's son, and yourself. How did that happen?

A That's a crazy group of people. It was, honestly, it was one of the most fascinating experiences of my life. I went to Yahoo to be the chief development officer, and once there I was asked to join the board of Alibaba. Because Yahoo was the largest shareholder. They had the most contentious relationship. If you remember, there was this dynamic where Carol Bartz had been in this contentious fight with Jack Ma that was widely chronicled, and the two companies stopped talking to each other, even though Yahoo owned 40% of Alibaba. And so I came in and was asked to help clean it up. And so what did that mean? There was like hedge funds and activists trying to force Yahoo to sell their stock, and agreements that were going to cost Yahoo tens of billions of dollars, and so I go into this situation having never been to China, and I show up in Hangzhou, and I met Joe and Jack, and Work to start to build a relationship to try to bridge these two fractured companies. I mean, this is like an unbelievable relationship, an unbelievable value capture for Yahoo and Yahoo shareholders, yet to not actually have a relationship with the CEO and founders of the other company is unprecedented, let alone the scale of the dollars we were talking about at the time. And so I went to China You know, as often as I could, multiple times in a month, I spent a lot of time with the principals there, and Joe a…

AI assessment note: “I went to Yahoo to be the chief development officer, and once there I was asked”

Q How does this then lead you to Square and Jack Dorsey?

A Jack's the best. So, you know, I was working at Yahoo, and out of the blue one day I get a phone call from Jack Who said, I've heard about you, and I understand you run HR at Yahoo. Can you explain that to me? I was like, I don't understand this phone call. It was so funny to just get a phone call and ask about my job, and so we talked multiple times about what it was like to run HR as an operating person and not a traditional HR person, and that was a model that hadn't been Use that much within Silicon Valley. And he was really curious about it because he thought that HR should have more of an operational bent to it than a lot of traditional HR leaders deployed. And I think he also trusted the function as an advisor. You know, at the time I called, when I ended up going there eventually, I felt like I was more of an advisor than a business advisor, than You know, what a traditional HR person would do, and so he asked me whether I would come to Square, and at the time it was before the Alibaba IPO, and I didn't think I could leave before the Alibaba IPO, and so I felt like my job wasn't done at Yahoo. So I waited for the Alibaba IPO. It was amazing. Then we reconnected, and I ended up Where I was in the middle of this huge tax transaction for Yahoo. There was a massive capital gains tax bill, and I was tasked with trying to solve how to manage that. And the complexity of that, …

AI assessment note: “out of the blue one day I get a phone call from Jack”

Q Wow. And all referrals. Can you list out some of your clients?

A It is all referrals, but, and honestly, I, our clients are really our sales team, and between our clients or me, um, that's sales. And so, you know, our first client was actually Ramp, and at the time, I was at a KOTU conference, and I was with Eric, and I pitched him on what we were building, and he was like, I'm in. And we ended up working on one product with them, and his philosophy was you can be no worse than the traditional banks that we use, and given the roadmap that you're going to build and the technologists that you're working with, um, you'll be incredible. And so I'm willing to wait the six months when you go build out your tech, uh, and advance your technology in such a way that We're willing to wait, and I know you won't be worse than what we work with, and so we got very lucky to have such a high quality client, and so, um, that was our first client, and then Atlas was one of our other clients, and they had just had a situation where another bank, um, ran into trouble, and Atlas had to come off their rails along with a handful of other clients, so they immediately came to us in an emergency and said, can you take us? We hadn't even closed the deal when I said, like, you have to go talk to the team before we own the company, but if they onboard you, we will absolutely deploy our technology to your products, and so we ended up onboarding Atlas as our second client…

AI assessment note: “our first client was actually Ramp... Atlas... Stripe, Walmart's OnePay, Affirm, Flex, Self”

Q of liquidity over the last year and will continue to be is secondaries, whether it's Anthropic, OpenAI, I don't know, Anduril, SpaceX, people that are doing tenders, it could also be Stripe. And so with those in mind, could you just talk through one, like the best practices for secondary sales and then going through these events too, um, and what to look for, what to not look out for?

A The secondary market is, um, you know, very, um, uh, active right now. Everyone wants to get, you know, a piece of the brand name, um, companies, right? And so this whole industry has sprung up around that. Um, brokers, um, friends of founders that happen to have a position But the challenge always with these things is that the companies generally are structured. They're sort of bylaws or structures so that the, uh, the employees cannot necessarily transfer the stock to outsiders, right? So what ends up happening is the employee creates a company. Molly Corp. And then Molly then transfers some of her stock into her own company. The company allow that because it's technically a transfer just to herself. And then she sells shares in that company and others. Right. So in these others, you know, people are trying to get into, into these companies don't really have a direct claim on a stock. They have a claim on Molly company. Right. And that's called the level one or L one type structure. And then what's happened is there's so much demand that let's say I bought shares in Molly Corp, right? I now go create Michelle Corp and sell shares in that to other people. And that's now a level two or L two, uh, type structure. So you see things L one, L two, you even see L three nested structures each time you're further and further away from the stock. And so there's more and more risks that…

AI assessment note: “each time you're further and further away from the stock. And so there's more and more risks”

Q Mhm. Okay. Well, what makes this round different than other rounds?

A I think, you know, this round one, we really didn't need the capital. Uh, we raised it for pricing purposes, and we saw a couple interesting opportunities to accelerate R&D, particularly around AIR, which is our LLM native EMR platform, um, and our voice agent platform. And so the idea is if we can accelerate timeline, bring in a team of 40, 50 killer engineers to attack some of these problems that are adjacent, but converging upon the same solution set that we have in RCM and ambient, it's gonna be net better for our customers. Uh, number two, we've extensively used CBF, which is General Catalyst's non-dilutive customer value fund to fund our go-to-market expansion without needing to dilute shareholders. And so the last two years have been very non-dilutive for the company and for shareholders. And this was, you know, a pretty non-dilutive round as well against the valuation. We raised not a ton of capital because we didn't really need it, but it sets a new mark and it lets us sprint and continue building.

AI assessment note: “this round one, we really didn't need the capital. Uh, we raised it for pricing”

Q Kabir is building the AI powered OS for healthcare. Can you unpack that a little bit more and the state of healthcare?

A Yeah, I think that's a great question. So, Camura at its core is a software company that builds tools for providers and for healthcare administrators. This is a multi-trillion dollar problem, and even when you cohort it down, there's a trillion dollars spent on healthcare admin every year. That's clinical admin, that's financial admin, like revenue cycle management, um, all of the tasks that happen in the back office of a health system, like, Scheduling, prior auth appeals. We've built a series of fine-tuned language models and agents that automate all of those tasks and truly take a health system that might operate at two, three percent operating margin today because of all the labor needs and turn it into something that can operate at 20, 30% operating margin, like the top percentile healthcare practices and systems. And our belief is that that's not going to be a hundred different point solutions. That's actually actually going to be one revenue engine. Um, and a series of agents that are orchestrated on a single platform, which is ComirOS. And so for us, the, the, the next couple, I mean, you know, today a practice that uses our platform will use us for their scribing, their practice management, their documentation, their autonomous coding, their revenue cycle, and all of their back office needs, and see that boost in margin because they can grow without needing to have lab…

AI assessment note: “We've built a series of fine-tuned language models and agents that automate all of those tasks”