If you read this newsletter regularly, you’ll know that Anduril has had a hell of a year.
In January, the company announced that it’s building a massive, cutting-edge production facility called Arsenal-1, and has since secured major DoD contracts for everything from cUAS, to an Integrated Visual Augmentation System (IVAS), to rocket motors.
And that’s just 2025. Since it was founded in 2017, Anduril has officially raised over $4.3B in venture funding and is now valued at $14B, according to Pitchbook data. (The company is also reportedly in talks to raise an additional $2.5B at a $28B valuation, but has not publicly confirmed this yet.)
In less than a decade, it’s grown from a scrappy defense tech start-up into, effectively, a new-age defense prime.
Anduril President and Chief Business Officer Matthew Steckman has been along for much of this wild ride. The Palantir and Zipline alum joined the company in 2018 and has overseen its super-ambitious business strategy ever since. Tectonic sat down with Matthew last week to talk about all things Anduril, how the company has been so very successful, and where it’s headed next.
Note: This interview has been edited for length and clarity.
Tectonic: What is the secret sauce that enables Anduril to be as successful as it is?
Matthew: That’s a big question. There is no one answer, so I’ll give you a bunch of answers.
To do anything new, to have any successful startup, you have to have a lot of variables working in your favor at the same time. That’s why most startups are not successful, percentage-wise. It’s hard to perceive from the outside looking in if you’ve never been in hyper-growth mode.
So, what’s the secret sauce? Well, we probably needed at least 20 different things to be correct in order for the company to work. Trying to simplify it is almost insulting to all of the people that work every day here and the complexity that we deal with. By the way, this has nothing to do with defense. This is just startups in general. Then, add in that we tried to do this as a hardware defense prime, and you add in yet another set of variables that have to be blowing your way in order for the company to work.
We find ourselves in a unique position where it’s 2025, and we’re being treated in the market and by the government as a true defense prime now. So, what were those 20 things that went right? I’ll just rattle off a couple.
The team has to be super switched on the technical side, on the business side, and on the financial side. We had an advantage in that regard in the early days. Anduril, for the most part, was not our first rodeo.
Second, in government and in defense, you have to have a deep understanding of how a defense works. How the DoD acquires things, how the branches of government work with one another, and how DC as an organism functions. You have to have a humble appreciation for the Rube Goldberg machine that is DC.
You also have to have a huge amount of capital to make any large hardware play, and you have to prove that that capital is being spent well in year-over-year revenue returns. Until your Series B, you can get away with a strong narrative, but after that, the financial returns have to be there.
You have to have brilliant technologists. When you enter the room as a business person—like I am—you have to feel like you have a tailwind behind you because you have the best technology team on the planet.
You have to be able to recruit and retain continuously and be able to do it at a scale that is really, really hard to even conceive of for a lot of institutions. We will be hiring over 1000 world-class hardware and software engineers this year. That is a tremendous feat of recruiting, which takes a lot of effort.
All of these things have to come together. And in defense, geopolitics have to be working as a tailwind for you as well. We shouldn’t forget that right now, defense and hard tech investing is in vogue. When we started Anudril in 2017, we were literally so nervous about starting the company in Silicon Valley that we put it in Southern California. I think that geopolitics has proved that we were right and that our thesis was correct. There’s a land war in Europe. There’s a threat growing in the Pacific. We all intuitively knew this, but it wasn’t well known in the broader tech community [in 2017]. It is today.
What has been the hardest thing about scaling the company?
That’s another one of those things where there’s no one answer. So, let’s just go through a bunch. We have every problem of a normal startup, plus all of the problems of trying to work with the Department of Defense.
Let’s start with the problems a normal startup has in a hyper-growth phase. It’s: How do we mature the technology fast enough? How do we find our first user group? How do we scale with that user group and understand what is working and what’s not? How do we operationally build the fundamentals of the company to find your second, third, fourth, and fifth users?
Then, in defense, you have added access problems. You have security problems. You have an acquisitions regime that can’t keep up with the expectations that the investment community has. In a traditional company, you would want a one-year cycle for ROI. In defense, it’s probably a 10-year cycle.
You have trust problems. For the largest addressable market in defense, you have to earn your way into being able to compete on those programs. You just can’t start and say, “I want to build the first-ever unmanned fighter jet.” That would be pretty ridiculous. You have to earn your seat at that table through lots of other work that shows the government they can trust your ability to deliver on its most important and sensitive programs.
All of those things are hard. Hardware businesses are hard because you don’t get a lot of cuts at the solution. Unlike a pure software business, where you can iterate a lot, in hardware, you can do a rev maybe once a year. Hardware is hard, which is why there aren’t a lot of new hardware companies.
All of these things are difficult. If any one of those things hadn’t happened, if we hadn’t gotten our first two programs of record within two or three years of founding the company, the company wouldn’t have worked. So, for all of the newer defense companies staring up the mountain, yeah, it’s a hard climb.
What drove the decision to make Anduril a hardware company?
We knew we were going to be a hardware business from the very start because we wanted to be a defense prime. Defense primes build large, complex hardware systems. The key insight, though, was that a lot of hardware programs struggle and go slow and then get mired in cost overruns and slips because the software that powers those physical devices lags or doesn’t work or is buggy. The software doesn’t work. So, we created a ubiquitous computing platform, and we called it Lattice.
The key insight was that we could create a repeatable software solution that solves a lot of the common problems that hardware in defense faces, like sensor calibration and infusion, commanding assets across long distances, resilient communication networks, or the ability to move objects in physical space, and do that in an autonomous way.
We set out to create, essentially, code blocks in a software platform where every new hardware problem we took on would benefit in some way from the work that had occurred previously. We could then take it like a set of lego blocks and apply it to the next problem.
So, so what did we do? We chose the easiest possible version of the sensor to effects problem, which is fixed-site awareness on the ground. Our first-ever product was for border security and military base security, where you just want to automate a function that normally you would have three shifts a day of three people doing. We automated away those nine people’s work by having cameras and radars and thermal sensors work and understand the environment around them automatically. To do this, we had to build a sensor fusion platform, we had to build a communications network, and we had to build the physical control aspects.
So then, what do you do next? Well, we had this pretty good compute system and hardware system that understood the ground. What is the next logical leap from that set of technologies? It’s counter-UAS. Can I take the same concept we used on the ground and point it up so I can understand what’s happening in the air?
We didn’t have to start addressing that problem with a blank screen. We already had a lot of the building blocks to do it. We just had to attack the new threats that we saw in the aerial environment. Now, in 2025, the crazy part about this is that the systems that drive the loyal wingman, our autonomous jet fighter with the Air Force, and the systems that drive the Ghost Shark, our autonomous underwater vehicle with the Australians, can all be traced back to that original compute platform. There are still code blocks that drive those large and very complex platforms today that are derivative of the original technology that we built as a company.
What’s your best advice for a defense tech company starting out now?
In defense, you have to think about the lily pads in front of you to get to the thing that you want to do. Even if you could build the best spaceship in the world that looks like the Enterprise from Star Trek, the government literally won’t let you do it unless you’ve taken the right lily pad steps to prove to them that you can do the smaller stuff. You have to prove you understand their problems and that you understand warfare. That you’re comfortable with concepts of survivability and lethality.
All of those are lily pads. All of those are separate contracts and probably multiple years-worth of work before the government will even allow you to bid on the Starship Enterprise. Think through that, be extremely rigorous about it, and don’t step on a lily pad that doesn’t ultimately help you get to build the Starship because it’s not worth your time.
The second thing is, I think that you have to realize that in defense, the only addressable market that matters to you is the really, really audacious and big one. It’s the starship. If you’re shooting at something lower, there’s more competition, there’s more access, and you’re probably fighting uphill against way more incumbents because it’s an easier part of the market to address. I always tell new companies starting out: If your end goal, the last jump on the lily pad, doesn’t scare the shit out of you, you are probably not shooting high enough.
That doesn’t mean you’re going to do it right away. That would be crazy. But it does mean you have to always hold it in the back of your head. For us, that final lily pad was that we want to be able to compete on every single major defense product and program that exists. That was our ultimate lily pad. I think that in 2025, we’re there. We would be credible bidding on effectively anything at this point. That is insanely audacious, for what it’s worth, but it took us hundreds and hundreds of jumps and lily pads to get to that point.
You guys have raised a truly insane amount of capital. When you’re working with investors, how do you convince them that it’s a good investment if they might not see a return for 10 years?
At this point, because we are so large, we are effectively treated like a pseudo-public company. When we go to the market and try and raise investment dollars, it is almost pure financial analysis. That’s because we’ve had so many successes where we can show that if you invest $1, you get $100 or $1,000 out on the other end. The return on investment is very, very good and significantly beats the general market.
Because we now have enough examples of that, and because we have enough new things that are entering the pot, we have shown the broader investment community that the returns are there. That now includes a lot of public institutional investors who have crossed over to invest in a private company.
This is important. The eight years of work get you to the point where you can actually tell that story. Where you have the knowledge and expertise to understand how those dollars will grow over time and how the government will think about those programs.
That’s why our recent funding rounds have happened so quickly. The story is there, and it’s just empirical at this point. I think that when you’re at an earlier stage, it’s about momentum, stickiness, customer churn, and financial data. But at this point, it’s a lot about trajectory. It’s “show me the money.” And we’ve been able to do that.
What’s the plan now? Where are you guys headed?
Similar themes around unmanned systems and autonomous systems. We’re focused on mass effects.
Take a very large, exquisite platform that exists today, like a naval frigate. It’s a huge, gray, old ship. Think about all of the missions that ship performs—hundreds of things. What we’re really interested in is asking, “What if I broke apart that frigate into 1000 separate robots? Could I effectively perform those 100 missions that the frigate was responsible for? Could I do it faster? Could I field it faster? Could I do it more effectively? Could the ultimate performance of those 1000 systems surpass the single exquisite system? Could I do it more survivably? Can I still perform the mission, and can I do it at a cost point that is 10x lower than that single exquisite system?”
That’s basically the framing for the future of Anduril. There are so many of these types of problems in different domains. It’s not just surface ships, but things in the air, under the water, and in space. How many of these types of problems can I tackle? We believe that for most exquisite defense systems, if you rethink the set of missions that those systems are responsible for in this way, you can pretty quickly convince yourself that the physics work in your favor. For the next five to ten years, that’s what you’re gonna see us do.
Do you think that, eventually, Anduril’s tech is going to make those big, exquisite systems irrelevant? If so, what does that mean for your relationship with the primes? Is that a tension point?
You need both. Will there ever be a world where you don’t? Maybe. Are we close enough to that world that I could confidently say that you don’t need the single exquisite system? No. I’m not worried about our relationships with the primes in that way at all. In fact, we have partnerships with all top five US primes, many European primes, and many smaller and mid-cap defense players.
We’ve become very good at figuring out where we have a value-add. As an example, can I take our software platform and level up the existing set of technologies that the US and our allies have? Can I eke out gains in performance? In the last 20 to 30 years, there have been trillions and trillions and trillions of acquisitions, and those don’t go away overnight. Pretending that they’re not there is not only disrespectful to the world that we live in, but it will also put a bad taste in the department if you don’t respect the decades of technology lineage that have come before you.
What we’ve become really good at is asking, “How much gain can I get out of the current systems?” The answer is a lot, and we’ve shown it. We just had a super interesting exercise that showed this. A team of software companies applied their systems to the Army’s existing kit to see what types of gains could be made for command and control.
The results were remarkable. I think that we’re going to have really, really strong partnerships with the existing defense companies and the existing technology providers for a very long time because it’s provable that with the smart application of, for example, new software, you can improve the Army’s ability to operate.
Obviously, with new hardware, the gains are tremendous. You should take advantage of that all day long. But when you map these two things together—legacy tech and newer systems—I think you end up being ultimately more successful than trying to do one or the other in isolation.
A pivot. Are you guys planning to go public?
We plan to become an enduring public company. That’s what we told investors, and that’s the plan.
Cool. Now we can pivot back. There’s so much going on at Anduril—from maritime to hypersonics this week alone—is that spread ever a problem? How do you guys avoid the pitfalls that primes have experienced as you continue to grow?
We run a complex business. We have about 20 different technologies that we go to market with. Most of our contracts feature multiple of those 20 coming together in unique ways.
The most interesting version of the answer to your question is: Is there a driving umbrella thing that helps us deliver faster at a better cost and a better performance? I think what that is, although I hate being reductionist, is the way that we structure the finances of the company. We are a defense prime, but if you look at how our profit and loss is structured, it looks like a high-tech company. You put a lot on your own books, you spend a lot of your own money in the form of internal research and development, and you get a product pretty far along. Only then do you go to market with it.
Customers obviously weigh in and help you complete the swing. They have their own unique requirements. But we take a lot on our own books, and we take a lot of risk with our own capital, and then when we go to market, we tend to fix the finances. In the defense world, we mostly use firm, fixed-price contracts. It aligns incentives between the company and the government so that we lose big if we don’t deliver on the cost schedule and performance as promised in the contract. The opposite is also true. If we crush it, we win big because we fixed the economics.
We’ve been really successful when working with government partners at balancing this sort of like risk-reward trade. They get a better product. They get it at a reduced cost point. They get it faster. We’re able to beat the typical margins of a defense company, then we use that to reinvest back in our own product lines. It’s a virtuous cycle.
At the defense prime level, that structure is unique. There aren’t many others doing that, particularly on the hardware side, where the risk you’re taking is tremendous. You see that bleed into everything we do, from how we build our products to how we structure our teams, to how we resource certain efforts, to how we talk to the government about what they’re going to see and what they’re going to experience. In our opinion, it’s the right way to do business. If we are successful, we should be rewarded, and if we’re not, we should be penalized.
So, if you were offered a cost-plus contract, would you take it?
There are definitely some things that you should do cost-plus. Things that are a complete unknown, where it’s entirely un-forecastable, unpredictable, and there is no way in your right mind that you would put that amount of risk on your own books. We are not dogmatic when it comes to these things. We have opinions, and we assert those opinions. But at the end of the day, we’re going to do the work in the way that the government believes they need to do the work. We’re going to try our hardest to convince them to do it in certain ways. But at the end of the day, we’re a defense prime. We’re just going to do the work.
We have cost plus contracts, we have cost accounting systems, we have all of the things that a defense prime has. There are some cases where when you look at how the program is structured, when you look at the finances involved, you can’t say, “I’m going to write a $500 million check on my books for completely unknown outcome.” That would be crazy.
Everything is a balance. Everything is unique. Everything has its own set of evaluative criteria, and you make the best decision that you possibly can at the time.
How do you guys, from a business perspective, decide what to build?
A couple of things. First, I think that part of the success of Anduril has been betting correctly and being a little bit or a lot bit ahead of everybody in making that choice.
In 2022, we bought Dive Technologies up in Quincy, Massachusetts. As a leadership team, we sensed that there was going to be a wave of interest, investment, and procurement in undersea systems, even though it wasn’t yet appearing within the federal budget. As a traditional defense company, you wouldn’t move a tremendous amount of capital around because the dollars weren’t there.
But through a lot of different conversations, both technical and non-technical, we were pretty sure that this was going to be a big macro movement in the US and allied markets. Because we’re structured as a tech company, we’re able to take risks. So, about two years ahead of those dollars appearing within the budgets, we acquired a small company called Dive, and we were able to spin that into an incredible partnership with the Australian government and an incredible partnership with the US government. We were ahead of everybody, which gave us a huge advantage.
So, if you trace a lot of the breakout technologies that we’ve had, it was mostly because of our ability to sense and our ability to move capital really, really quickly so we can get ahead. A bunch of our products can be defined by that set of criteria lining up.
When you say sense, what do you mean? Talking to people in the Pentagon?
We knew when we started the company that we needed an apparatus that just understood the Pentagon and understood the services and understood the Congress and the White House and how all of those entities interplay. That’s a super hard capability to build. It’s also really expensive. This is a huge barrier to entry in defense. The defense primes have had 100 years of building that capability. For new companies starting out, you have to build it from scratch. So, a big part of what we did in the early days of the company was building that to get ahead of the market.
When you’re looking to acquire companies like Dive, what are the green flags you look for? And what are the major red flags?
All of our acquisitions are because we thought the technology, in some way, was groundbreaking. There are a lot of acquisitions in the market where you buy to margin optimize or vertically integrate. We don’t do that. We buy technology that we think is incredible, and we like to acquire and work with companies that we think have incredible engineers. If you look at all of our acquisitions, that defines all of them.
That’s criteria number one. Criteria number two is: Can this company or technology instantly gain from the current technology portfolio that Anduril holds? If I smartly apply our Lattice software system to it, is there an instant unlock?
Criteria three is: Do we, as a company, have a pretty darn good idea of how we can go to market with the tech? Can we use the apparatus that we’ve created, that DC sensing apparatus, to supercharge the business that we’re now merging into ours? I think we’ve done that, by any measure, really, really successfully. For investors, they get 100% gains year-over-year for many of these acquired companies. It’s insane. When you see those three criteria line up, it’s an easy decision for us.
What’s your perspective on overvaluation in defense tech?
Maybe it’s just because I’ve been doing this for 20 years, but I don’t think this is unusual in venture investment. There’s a reason the power law curve exists. You have a thematic set of investing that happens really, really rapidly with a tremendous amount of capital. And there will be one, two, maybe three, power law winners off of that technology trend. So, the valuations are not wrong if, ultimately, a handful of companies go public and are valued by public investors and institutions at that value.
What does that really mean, though? It means that for the rest of the power curve, all the other dynamics of the traditional venture market play out. Companies get acquired. Companies are devalued down to non-venture or non-growth. Many companies will go out of business.
You’re not going to have 100 winners or 50 winners, you probably won’t even have 10 winners. You might have five. And that’s okay because the interesting part about defense and national security is that after this 10 to 15-year period of investment, you will have five new entities that can conceivably bid on complex programs. You have changed defense at that point. You have dramatically moved the needle. You have innovated. You have increased the deterrent function of the US, which is why we all do the thing that we do.
What is the biggest mistake that you see smaller companies making as they try to enter the industry?
Every single technology area in defense is effectively a moonshot at a monopoly. If you’re starting a company in a technology area, there’s only one, sometimes two, contracts that you have to win in order to have a viable business. If you don’t win them in that technology area, you don’t have a viable business.
It’s very unlike the commercial market. It’s very unlike enterprise sales, where you get a lot of shots to be successful. In defense, if you take your shot on a goal and you miss, you’re going to hurt.
So, what does that mean? It means that a business that can support multiple types of things will be better hedged. Obviously, businesses that win the monopoly will do well, but it also changes how you think about your first few years. This is starting to become more understood, but I still find that I’ll talk to a new a new startup founder, and I have to tell them that there’s literally just one contract in the Army they have to win, where the lion’s share of the budget will go for this type of technology. If you don’t know that ahead of time, then you’re going to have a really steep learning curve.
What are Anduril’s overseas plans? Are they impacted by tariffs?
We are already in the UK and Europe. We’re already in the Middle East. We’re already in Australia. We just opened shops up in Japan and Korea. We’ve really, in the last two years, become a global defense prime with direct contracts with allied governments and through foreign military sales, sponsored by the US government, to those countries.
The plan is to continue to grow and expand internationally in a lot of ways. We’re advantaged by the type of thing that we build in those markets. Because the products are finished and non developmental, and they tend to be less expensive but still performant as compared to need, we actually tend to do pretty well when you butt that up against a lot of what the US and European primes are making and building. Our outlook internationally is quite strong in terms of our forecasts and in terms of the level of investment and capital we’re putting into making our international businesses work. We feel pretty strongly about it.
Tariffs will impact every company. There’s a lot of nuanced, weird, technocratic reasons why defense is a little bit insulated from that, and we’ll deal with it just like every other company does. There’s no real difference for us there. Our finance people will go to work, our supply chain people will go to work, and we’ll figure it out.