Fireside Chat with Brian Schimpf, co-founder of Anduril

Thank you.

I wandered around campus. I hit College Town Bagels because got to have some nostalgia to it. Unfortunately, it had moved across the street for when I was here. So, it was like on the corner right over there. But it's much nicer now. And then just wandered through North Campus. Are the freshmen still on North Campus? Is that still the deal? All right. I think it was just starting when I came through. So, yeah, did that and then did the loop. It was great.

They look very nice. Some of them are very nice.

So, I grew up in Rochester, New York. So pretty close by. And I've been like coding since I've been 12. Like I'm just a programmer at heart. And then when I was looking at different schools, I visited like five or six, but my parents had actually gone to Cornell and so I'd never really like been because of that. I was like, eh, I don't know. But actually when I visited I was like, ah, this is great. Like this is the spot for me. It just seemed to really resonate well. So I just applied early admission or decision or whatever, got in and I was like, all right, well that's done. Great. Easy. So it was like super straightforward. And I don't know, that's just been a pattern for me. I just decided to do a thing and it just go.

It's Henrietta's the town. So, one of the suburbs there.

Yep. This was before Palantir. It was only four years old at the time, right? Yeah. Yeah. I think I was like employee like 70 or something like that.

Yeah. Yeah.

Yeah. Karp likes the mysterious vibe. Yep.

No. It was really just Karp of that founding crew. Lonsdale left. Peter is not an operational guy. He is a big thinker, let's say. And so he was, I don't know if he was ever really operating with it. And then yeah, I was one of the early hires on the forward-deployed engineering team, which was really just about how do we get out at the customer, implement the product, get it working, figure out what the customers really need and feed that back into the product. Ended up leading that on the engineering side and, but yeah, was one of the first handful of folks there. The other guy who was running it at the time, also a Cornell alum, this guy Sham Sankar, who's awesome. He's now the CTO. And, yeah. So, it was pretty cool.

You should get Sham. Sham would totally come back. He's a hoot.

Sham Sankar.

Yeah.

Yeah. Only interviewed at Palantir. I had a friend who I'd worked with on this hybrid electric vehicle student project team and when I was here the student project teams weren't like a formalized thing really. It wasn't like, it was like two and by the time I left I think there was like five.

Now there's like 30 and it was like space and the university doesn't want to kill you, it's like great. And so yeah, I had a good friend who I worked with on that and then he called me up was like, 'Hey, you should take a look at this Palantir thing.' And went checked it out. 2007 it was like, it was very different. Like entrepreneurship was just not part of a university culture. And also for if you were on the East Coast 2007 it was very hard to get into that Silicon Valley startup scene. Like it was just not a thing. Like all the advice was like go somewhere safe, go to Lockheed or a bank or Microsoft. Even Google was a little too risky 2007. Like you were just really discouraged from doing anything kind of outside the norm on this. And so yeah, my buddy was like, 'Hey, we're doing a career fair. You should check it out.' And I was like, 'Oh, this is cool. Like I want to go do this.' And yeah, just was like, why not there? You were based in where? Where were you based?

Okay. So Palo Alto. Gotcha. And you when you started you weren't the director of engineering, but you eventually got to that spot. Yeah. Can you tell us a little bit about your ascension?

Yeah. So I originally started working for the dude that recruited me. He turns out he didn't like management, so he's like, 'Hey, you want to swap?' And I was like, 'Sure, that sounds good.' So I ended up leading that deployed engineering stuff working for Sham and then we just kept growing in the different areas and you know, I kind of had an aptitude for the management side, the strategy side, like building the org, all of those things and so I just kind of figured that out, learned it. And this progression of like, it's almost like weird if you haven't gone through this hyperscale growth thing, like it's just very hard to appreciate what that feels like but once you're used to it it's all quite predictable. Right, you're like okay, like every factor of two scale, everything you assumed was working no longer works. Right, you double, everything breaks. All right, fine, just plan for that. Right, like you're just constantly in this mode of change, of reorganizing, figuring out a new way to do things, changing your hiring profile, like everything's just constantly morphing and shifting. And for a lot of people, that's quite uncomfortable, right? Like you can't just scale the one thing you were doing and it just works out. This idea that your organization every year will look totally different and everyone you hire has to be very okay with you're just going to figure out what your job is when you show up. And even at large scale, that's still true. And if you build a culture around that, it kind of works. And so I just happened to be quite good at that. And then the head of product left. So then I took over kind of product as well working for Karp. And it was an interesting time in the Palantir journey where we had these very defined products that were working in the government side. The first attempt at a commercial product just totally failed. Like it was not the correct thing for the market. It was very cool, but it was a very niche product that only worked for like two hedge funds. And so we had to refigure out something new. And it was a bit of a weird experience because it was like the first time in a proper product executive role where I remember going to Karp like, 'Hey, we should talk through this product strategy stuff.' He's like, 'Dude, I'm not a product guy. You figure it out. That's what your job is.' And I was like, 'I guess that's true.' So it was just had to assert like, 'Hey, we're going to go do X now and get everyone bought in on it. Build it.' We built this thing called Foundry. Rebuild the entire engineering team who had kind of atrophied to an extent because they were like kind of in this sustainment mode. Like rebuilding an organization that's got to change dramatically. I mean, it's a much more invasive process than I think people imagine. It's got to replace all of middle management. Like, it just took a lot of work. It was a really interesting learning experience. And then just putting forward like here's what we're going to do. I'm convinced it's right when it was just kind of like I just got to do it. Just go. And I had a ton of support from Karp and others on just going and doing it. But it was just a really unique experience. Just go build something from scratch and make it happen. And yeah, so I did that for probably about three or four years before starting Anduril.

So, the government side we kind of hit it right early, but it was a very... well, the initial product hit pretty well and then there's several subsequent pivots, but this for deployed engineering side was really, really valuable. I mean, I think Palantir figured out a lot of the right things on how to do kind of this enterprise software model very, very effectively. It's very different than like a Snowflake or these others where like the typical sort of enterprise SaaS thing is like here's a fairly boxed solution. You're going for large scale, but like the ticket size per customer isn't like that huge. Like any one account isn't that significant a percentage. Palantir really dialed it around a very white-glove service, disintermediated all the Deloittes and all these other companies that were like, you know, in the past when you're doing a lot of data analytics you bought some ridiculous software from Oracle plus then a massive team from Deloitte to implement your data warehouse strategy and they took 90% of the revenue. And like it was just not very good. And so having this model where we just had embedded teams with the customers, it only was like five or six people, they were never large teams but you get like five or six very hungry, very empowered young people. They're just very talented. They will do an immense amount of work and thoroughly enjoy it. And for the right crew, it's like going hanging out in Toulouse for 6 months and working with Airbus. That was a pretty sweet gig, right, for like the right character of people. And so we just got this model where we were just able to constantly deploy these early stage products, figure out if they worked or not, get them dialed, take these pilot customers, repeat. The model that doesn't work that a lot of people think is true is that like this innovation bubbles up from the bottom and you're just harvesting is kind of not true. Like the version I think is true is like the strategy to make a good product, to make these things actually work, it requires strategic thought and real intention of where this is going to go over a long period of time. You are probably not going to arrive at that conclusion through some random entropy that you're harvesting, right? You can get ideas, you can get things, but it's probably not just some random prototype software someone wrote that you're just scaling. It doesn't work that way. In reality, I had to have this very intentional process of like how am I building this to actually be scalable and then I use the fact that we have this access to say okay these customers would prioritize these things, these are the right beta customers, let's trial it, get it dialed, repeat. So you could just cram through adoption and take beta quality products, get it out, band-aid it fast, learn much faster. So you just close that loop way, way faster and that's if anything on the software side that has been the critical thing over the last 20 years, right, has just been how do you get feedback earlier and often and lower your cost of mistakes. It's just lower the cost of iteration, increase your velocity. You need to think, but you can be a lot less right, like you can just iterate your way through a lot better. Like you have to have intentionality, but you can learn a lot faster. And so that's the whole game with most of these software companies is just how fast can you learn? Like how fast can you get data and feedback on these things?

Define visionary of the company. I love Alex. Like, he's not a product guy. He's not an operator, but he's a sales guy. And he has this incredible sales and talent guy. And so I would say the things he focused on were number one creating a space for like highly talented people to thrive. And the reality is like highly talented people can be quite annoying. And the management apparatus that gets built at any company tends to value compliance with the management apparatus, not actually doing the right thing. Right? And so he created a culture that was valuing these highly bright but highly annoying people, right? And like there was this one guy who was brilliant that I had working for me and I think Karp described him as 99% right and 99% annoying and it was like pretty accurate. And so like but he created this culture that really valued the brilliance, the artistry of these things and that worked out very, very well. That's probably number one.

Oh, it's total chaos all the time. You just got to learn to live in the chaos. Just doing it. It's lovely. It's okay. It's like, you know, engineers don't do what you tell them to do anyway, so like why would you pretend they are?

I have a probably a similar philosophy around it is a managed chaos. When you're doing hardware, it's a little bit different. It turns out like this whole goal is software of okay, I can lower the cost of the stakes and increase velocity. You can't do that with an airplane, right? It's just not how it works. And you can do it with cheap little airplanes. Like you can crash a lot of little drones. That's the right strategy. You do it with like a $10 million airplane, it's like you can't just put that many in the dirt. And so, and the cost of mistake, the cost of not thinking it through, compounds when you have these more expensive, more difficult things. And so, it's required a lot more structure, but then the structure... so there's a question of these organizations like how do you think about structure, which is like the big companies tend to value compliance with a process and no deviations, right? Like that's the thing they sort of value and there's a belief that like our process and how we do things is so valuable that it will continue being valuable in the future. So the tendency for most of these large companies is they make X software they keep selling it, they have this Y hardware process that will work now and forever. And so we're... it's about refining that into a process then compliance with the process becomes very important. This is not how we operate. We say like, 'I don't care how you do this. You have to make an airplane.' And we're going to empower you to do it. You can hire who you need to. You can do what you need to do. I put some financial guardrails on you and things like that. Some safety guardrails. It's like if you're playing with explosives. Guess what? There's a team checking your work.

Yeah. Like we share, we create a lot of cross-sharing and like... but I just think these things are like very overrated and the right strategy is highly autonomous teams. It scales very well. You get much more empowerment and buy-in. And this was the other thing that was like the magic with Palantir was all about creating a ton of autonomy and ownership. And if you can construct that ownership, you get people just crawling through broken glass to make things happen. If you don't have the ownership, it's just a job.

Yeah. It just doesn't matter.

Yeah. I agree. And then the other thing I find is getting people from these big cultures, big company cultures. They tend to focus like this process orientation shows up and they'll focus on like we had this plan and look at all the work we did. And I'm like, I do not want to hear about what work you did. Like I'm totally uninterested in that. Like did it work? Like what was the results? Like how are we progressing through results and setting up these high performance kind of cultures around it is just like accountability to results and like that's the only thing that matters and how you get the work done I don't care after that. Like it's just do what's efficient.

No, so it was interesting. So two of my best friends, one Cornell guy, Matt Graham, that I spent a bunch of time with here. He was at Palantir with me for a while. So the guy Trey Stevens, so we all worked together for a while. Trey and Matt went into venture and then Trey was at Founders Fund and he was like, 'Hey, I'm looking for a defense company to invest in. One doesn't exist. We should just go do it.' So we probably talked about it for like two or three years about starting Anduril. And then, Founders Fund invested in Oculus. They run this retreat every year. And...

Oh, Oculus is like, you know, the VR headset. The... what do they do? They still call it Oculus?

Yeah. Yeah. Yeah. So, he had sold the Oculus VR thing to Meta or it was Facebook at the time. And then got fired for basically supporting Trump. And yeah, and so he got out, but Trey had gotten to know him through Founders Fund. And then was like, he wanted to do the same thing. So once he got out of Facebook he was like, all right, let's go.

Just different. And so you know with Anduril they were like, hey, would you do CEO role and I was like that sounds cool. So when we were getting a team together and you know I think it was like they did the... I was like originally was like head of software and I was like yeah I'll do it but like couple years I want to finish some stuff out of Palantir. Then it was like they did the first pitch meeting, didn't have a CEO and they were like this sounds cool you can't have a company without someone in charge what are you doing that night. Trey and Matt called me and said hey would you do CEO. I was like, 'Oh, that sounds good.' Hung out with him Friday, talked to my wife, flew down to see Palmer Saturday, and then Sunday told Karp, 'Hey, I'm going to go do this thing.' And so the Palmer thing is very different. Like, so one, I'm in the CEO role, so I'm just running the stuff. That's different. Palmer is just a... they're very different people, but I would say they're like Palmer has the things he cares about that he wants to stay engaged in and he likes doing, you know, like media stuff. It's like great, works out great. He's extremely good at it. It's awesome. And so like we just have this great relationship and he cares about the things he cares about, right? So, he is an incredible inventor. He is amazingly good at like the things he really wants to dive in on. It's like we're working on this cool soldier-borne augmented reality stuff. It's awesome. Palmer is super passionate about that. But he doesn't want to manage. It's great. So like it works out really, really well. And then we've just created a founding team that's just very high trust, can just work through problems. We don't really have like disagreements. Everyone's got like different stuff they work on and it's pretty straightforward. Yeah.

Yeah. Yeah. Because of a bunch of different things. So the number one is like understanding what the sort of hypergrowth looks like of like how do you scale a company through thousands of people is just it's just a weird thing that if you kind of haven't gone through it and really understanding that talent piece and all those pieces are really, really valuable. The second part was because of where I sat in the organization and the different things I had done, just huge amount of exposure to the business and sales side which like as an engineer you don't typically get but like really understanding like how do deals actually happen, like how does the sales process, what does even a good functioning sales organization look like. Like that was just like a learning experience going through with Palantir. What does a recruiting organization look like? Like what are all these ancillary functions you need to have? There's tons of exposure to that and then I think like the product strategy piece was really, really valuable as well.

Yeah, it worked out.

Yeah. I just think like my bias has always been just like I don't really worry about like the job title or whatever and like kind of as I was moving through stuff, it was always just sort of know what's really important and what really matters to the company and just dig in and do it and just like get on the hardest most important things you can and win. Like it tends to work out pretty well for you. And that's just sort of my bias on these things. So, I think like if you've got the aptitude and you just go and you just don't worry too much about the credit or what's my title or any of these things, like it tends to just work out. So, you know, I'm not actually sure how path dependent it was. I actually kind of think like these things tend to be like the macro tends to work out. It's like, you know, if you're just driven, you're going to go after hard stuff and you're just going to do it and you got the aptitude. Yeah. There's a degree of luck of like are you working on the right problem at the right time like the market timing dimension. But the really good people tend to do really well in the long run.

Like, this is... if this is right for you then I support you and you know it was emotional like I've been there 10 years but he's just an absolute class act. So no, no drama at all.

A lot of practice and then I had a couple of good mentors along the way. So it's like starting out, you know, I kind of see this progression with most people moving through different kind of tiers of management. You know, you start out and it's like, how do I get a small group of engineers, let's say, to accomplish a project, right? And it's like task management. How do I divide up a project? How do I structure those things? And with that, you're kind of learning then you move into more like career management. How do I motivate people to understand and grow people to think about their potential and what they're trying to do? And I think that gets very overwrought. Like it's like all these frameworks, all these things like I'm just kind of suspicious of it. The only thing I've ever found for high talent people is you give them really hard ambitious stuff to do and a lot of ownership over it and they tend to do really well. And it's like a little... This is true for engineering, this is true for sales, it's true for a lot of the white collar jobs. Blue collar is a bit of a different situation. And then once you move up to more of the executive levels, it's more like how do you lay out strategy? How do you motivate teams of teams to get things done? How do you incentivize them to get the right outputs? Where do you need to drill down and drill into things? Where do you need to just set up the right frameworks and organizational structures and incentives to get the output of it? You're moving into more like org building and how do you as an executive spend your time to get the right outputs. And that's a very different skill and then you're kind of in then strategy and whatever else. And so kind of got this progression as you move through it. And it's just a toolkit. Like honestly what I've always found is it's just like just try things, find what works for your style. At first you're going to be like, 'Oh, this person's not performing. It's me.' And then after you do it enough times, you're like, 'Maybe it's me, maybe it's them, but guess what? I'm in charge.' And like this is... they're either going to work or they're not in this setup. Yeah. And then you just get enough confidence that you're like, look, I don't have to be perfect. I just have to win. And some people will win, some won't. That's okay. Like this is the style that works for me. And you just get enough confidence over time on these things.

No. Read a couple of books that people were sort of prescribed reading at Palantir and I didn't replicate the process at Anduril. Bizarrely the one that stuck with me the most was a couple of chapters in this book called Improv. And it was about... so it was this fascinating book that was like kind of explicated... it walked you through how to perform to appear to be performing a certain way, right? It was like acting, right? And it was just sort of like so one was like on status and like how do you convey like, you know, being more alpha versus beta? How do you think about that in the room and all these things and it's like that's the only one that stuck with me. So, it's a terrible book. I wouldn't recommend reading it, but there was one chapter that I remember. And then it's like there's a lot of things like it's hard for engineers to figure out how to do the hard conversations. Some people are good at it, some aren't. I don't know. It's just like you just got to figure out what works for you on these things. But like most of the books are not super helpful. Gives you some ideas, but I wouldn't say you'd walk away saying like I understand management tactics. It's just you got to practice, you got to try things and you got to be challenged in these situations.

Yeah.

Corey.

And we're going to move.

Yeah.

Yeah.

Yeah. Yeah. Yeah.

Yeah.

Yeah. Yeah. She did well, too.

Yeah. In an industry that nobody loves except for people like Peter Thiel.

No. No. They like literally weren't allowed to. They had to change their LP agreements. Yeah.

So for me it was like founding team. Yeah. Ambitiousness of the idea and the ability to like line up capital based on largely the credibility of the founding team. It just like it was just sort of like so teed up to be a monster that it's like how many potential things are you going to get like that. So it's sort of like if I'm going to do it, this is the thing to do it on. And it's sort of just like I don't know. It's just sort of like everything else I've done, I'm just sort of like, yeah, seems like a good idea. Let's go.

Yeah, I like... I'd done well off Palantir. I'm like I was not going to hurt in any scenarios. I was going to do great. But I think this risk thing is like massively overwrought, right? It's like okay, so say one of you guys in here start a company and it doesn't work out. How long do you think you will be unemployed? Is it measured in minutes or hours? Right. It's like what are we talking about? Like so this risk thing when you...

What are you talking about? Right. Like it's just it's an ego risk. It is not a financial risk.

Yeah. Well, what big tech does is very smart. They get you addicted to the lifestyle, man. It's like they're like, 'Oh, here's your fat 800k a year salary.' when you're like middle, you know, like and you're like, so you're just in the lifestyle like so the risk is your lifestyle. Um, and it's like not when you're 25. It's great. Um, so like I think this risk thing is just you also take a lot of risk.

Probably more but yeah something like that close. Um big valuation currently around 30 billion at least that was the last race probably more now um founders fund and drees general catalyst fidelity sans like it's a lot of big egos.

So um uh the investor side is very simple. Is Mark involved with Andre? Mark's involved um there is investor beef they are competitive and that's just kind of how it works right so it's So the investor the venture dynamic in so for the tier one venture funds it is like the only thing that matters is deal flow and allocation in the top companies and and actually these venture funds more think of their the way they market their success is not their IR like that matters uh when someone's doing diligence but the marketing is the tier one companies they are invested in. So deal access to tier one companies is uh is is the most critical factor. But then the great side is once you're a tier one company, you have unlimited investor demand. And uh and so the um and then like I have a bunch of friends doing startups and they're always sort of like oh should I go adventure thing? Why or this? And these guys are saying they can do all these things to help. It's like, dude, if the investors were like the reason you were going to succeed, like you were probably not going to succeed, right? It's like their help. If like you couldn't figure it out, you weren't going to figure it out, right? So, they could be good mentors, advisers, good like to bounce ideas off of. Some can have useful political connections. Like that can be valuable, but like fundamentally it's on you as a startup founder to like win. Like it just like if if it's just how it works. So like I think that's overrated. I think the the mar the to become a tier one company you need tier one investors. It's this very circular thing that is like very much around brand and creating this perception of momentum and exclusivity. And that worked out very very well for us. Um and um and so the most emotional things with investors is just it's just allocation. It's just allocation and cutting them back and all these things and who gets in and who's out and that that I've had some very aggressive conversations with different folks around that, but it's that's a champagne problem. I'm I'm not that upset about it. So, um the uh uh so like that's that's been really it. And the other part of this I'll say is if your company's doing well, no investor is going to give you a hard time. And if they are, get rid of them. um if you're doing poorly, every investor is going to give you a hard time and they're not they're like, 'Oh, we'll be here for you for the hard times.' They won't. Uh so like it's just it's just not how it works.

Well, our board is officially B train Palmer. We've had give up zero board seats. Yeah. Yeah. It's great. Again, if you're doing really well, like you can do whatever you want.

We were a very high potential startup with a great founding team.

We have board meetings and then we now have observers. So a bunch of the investors are observers. Yeah. They come, we hang out. It's a good time. They can't vote on anything. Um but uh

Yeah. Without the annoying venture people. It's really great. Uh so the

Yeah. It's very good. I think the board controls this.

It's very rare to pull off what we've pulled off. Um, again, it just worked because the demand was high enough that we're like, so we just negotiated away the board seats. Well, like no one no one would like in the tier one guys don't even ask for board seats really or like some do but we're like no and they're like oh okay that's fine. We're just like we don't budge on that and they're like great. And so we've just kept clean investor rights agreements, clean board structure, clean control and then it's like squeaky clean. It's just super clean, super easy.

Yeah, so we've started doing so with the observer thing. We'll go through financials, right? And we started like we're not in any rush to go public. We have no particular desire to to rush into that. We're just in a position we have basically unlimited private capital and a whole bunch of reasons. You just don't need to do it. Um and uh but the good parts are running a financially disciplined company, right? So like you actually want to run a company that has forecastability of cost and revenue. That's a good thing, right? Like allows you to plan, allows you to know that what you thought was going to happen happened and you can build that machine up. Um like you want to run efficiency in terms of like cost and showing you're getting more efficient on utilization of all these just there's all these things you want to do to run an efficient well functioning business. Um, so we do that. We run quarterly, you know, we run our audits. We do all that stuff. Like we're doing all the things you would do as a public company that I think are the healthy things to do as a public company. Um, but we're not in a rush to go public. So then we do report, you know, quarterly, annual earnings, like we have a financial plan.

Yeah. We have total We also have like 50 to1 super voting on the founder shares and like it's just

Yeah, we've just locked it all down at every level. So, it's very good. Yeah, we're very paranoid about this. This is the other carp lesson was uh control of the board is critical and do not mess around and um so we just from day one we were like found the right corporate lawyers who got it. We interviewed like probably 20.