# SpaceX, Palantir, Anduril & Onodrim | The Return of Conglomerates Sourcery · 2026-11-13 · 61 min · https://www.youtube.com/watch?v=4J7tjyVCrqA ## Transcript Molly O'Shea This year has really been the year of SpaceX. Christian Garrett It's a company we've invested in for 16 years. It's our largest position, and it grew its top line 160%. They posted $4.4 billion in net income and $8.8 billion of EBITDA. It's a $2 trillion company that is a dominant provider in multiple critical industries. It's clearly a very exciting company. Just a handful of companies drive all the returns. That's obviously how we've built the firm to capitalize on that. Molly O'Shea Cognition is recently at $48 billion, and Anduril at $90 billion. Christian Garrett I would just like to return to the conglomerate thesis right now. China’s defense primes: 60%, 80% of their revenue comes from the commercial sector. Palantir and SpaceX are the only examples of this, but I think we need to return to the dual-use conglomerates. General Mills made cereal and weapon systems. Molly O'Shea Christian Garrett, welcome back to Sourcery. Christian Garrett It's good to be here. Molly O'Shea I know we're in this lovely spot in a secret location, as always. It's great to have you here. We have so much to talk about. I know I've talked about you, or talked with you, a bunch over the last year or so, whether at Hill and Valley or on the podcast. We've talked about SpaceX and Anduril and vertical integration and all these things, but this year has really been the year of SpaceX. I'd love to hear from you how that's evolving and how the business model is forming now that it's a public company. ### SpaceX Becomes a Hyperscaler Christian Garrett Yeah. It's a company we've obviously been invested in for 16 years, and it's our largest position. It was a great win, not just for us, but for many folks from Sequoia, Vy Capital, Founders Fund, Valor, and others. A lot of people in the ecosystem are very happy. The business is honestly just getting started. They just had their Q3 earnings, and as you saw, there was a huge acceleration in the business. They grew their top line 160%. They posted about $4.4 billion in net income and $8.8 billion in EBITDA. It's pretty wild, given the growth of the business. The core business is still growing. You look at Starlink, and that's moving way faster, particularly on the D2C side. Then, on the compute and AI side, whether it's Cursor or Grok chatbots, they're showing the ability to productize and monetize their AI products. Going down the compute stack, they've been able to build infrastructure and use that capacity for themselves or sell it. I think the thing that's been most interesting, and is underappreciated, is that the company is clearly going the route of a hyperscaler. They're building out more software and services around that compute stack and new infrastructure. They're committed to this business. This is not just a one-off thing. They are the best in the world at building this capacity and building it the fastest. I think that gives you more durability and visibility into this business. The second piece I think is most interesting is the fact that they reiterated continually on earnings calls that they're getting sub-12-month payback periods, which is obviously a signal for demand. It's hard to forecast demand long term. This is what the markets are struggling with if you look four or five years out. But right now, there clearly is not enough supply to match demand. The other fascinating piece is that if you're going to invest anywhere within this space, you probably want to own the vertically integrated players, which they are. They can monetize at the app layer, the model layer, and all the way down to compute. That gives you some risk-adjusted investment exposure because whether there's a shift towards open source that changes the economics of inference and compute, or closed-source models continue to win out, they benefit either way. The hyperscalers, as we've seen with Google Cloud or AWS, have seen huge growth regardless of the shift in the volume mix of models. You also have the ability to shift capacity towards your own products if demand slows down in the outer years. This demand's not going away, so it's not a deterioration of the core business. It just changes the growth forecast and the CapEx risk, but I think there's nominal risk if you can use the capacity for yourself. If you build these things fast, you can pull back on them really fast. Your CapEx investments are in shorter-duration cycles, which means you're going to get a payback period on the outlay, and if the markets shift, you can shift quickly. I think they're clearly running leaps and bounds around the neo-cloud market and stepping into being a true hyperscaler. That was particularly exciting. Then, obviously, you continue to see more investment in scaling out Starship, which is kind of the core business around launch. Starlink is not just compounding, but moving faster into D2C. You see a pretty impressive business. I think Elon's forecasts may be very aggressive, but he's always been right, depending on what timeline you pick and what timeline plays out. If you look at this business now, consensus EBITDA based on Q3 puts it anywhere from 20 to 30 times. Twenty-seven times EBITDA for a $2 trillion company that is a dominant provider in multiple critical industries—it's clearly a very exciting company. Molly O'Shea I have so many different paths going through my mind right here. Number one is, the story about SpaceX from the beginning was launch, space, getting to Mars, getting to the Moon, increasing launch capacity, that kind of thing. What you're mentioning is this dramatic shift to almost becoming a hyperscaler, if not already one. Even in their earnings or their S-1, the story was that they're an AI company. In their earnings, it was reported that over 50% of their revenue is attributed to AI in some capacity. What do you make of that? And secondary to that question, what would their market comps be now, and what will they be? ### SpaceX’s Many Growth Engines Christian Garrett It's sort of an ode to the great companies of the past that I've become more and more obsessed with. You look back at these large conglomerate-type businesses in the '80s. There were always benefits. There are negatives to those businesses, which I don't think are attributable to the next-generation ones like SpaceX, and we can dive into that. But one benefit was that you had multiple business lines exposed to different industries that at times would provide diversification. If an industry cyclically was growing much faster than another one, then you at the top-company level could continue compounding. Here you're seeing one business line with exponential demand, growing extremely fast from a small base compared to other parts of the business, whether on the connectivity side or the launch side, that are in the beginning of another S-curve in our view. Those just play out over different time horizons. Yes, right now the AI business is exploding and moving very fast, and they're carving out a huge part of that market for themselves. But if you look at launch and Starlink in particular, I think what's underappreciated once again is that Starship is working. They actually got it launched and had payloads delivered to space that are delivering compute on the V3 satellites for Starlink. The next generation of the constellation is being built out, and people really don't understand the scale. That launch, with just a handful of Starlink V3s, delivered about 1% of the world's internet capacity. It's insane, and they're going to deploy tens of thousands of these. Then, obviously, there's the D2C side. Over time, you'll see this revenue mix adjust and shift around as these other businesses grow. With Starship as a platform, it's unclear how much the core launch business grows. I think it could be much, much larger than it is today and a much larger percentage of revenue. Starlink is similar, with D2C, V3s, and various other products they're continuing to build. I think people generally don't appreciate that these are all just moments in time. This is a business taking a multi-decade or century view, and I think the revenue mix will shift and adjust as it does for all businesses. Every business in its early days—which is hard to fathom for a 20-year-old business—is in its early days. You'll see this kind of revenue split shift based on demand across their different industries, particularly when you have a company with exposure to multiple industries that are in their own different growth curves. Molly O'Shea Well, it's interesting because everybody's in the early days now. Everyone's trying to grab on to inference revenue or compute revenue, or all this kind of stuff, and build out their business models. We're seeing this, and I don't know if this has led to delays in IPOs or whatever, but people are trying to figure out: okay, we have a lot of this growth. How do we backfill this valuation if we just have one line of revenue? Meanwhile, SpaceX is an incredibly durable company. It has 10 or whatever—how many do they have?—different ways to pull on different strings in case one thing falls or they need to lean into something else for more momentum. We were just at Starcloud in Seattle. Philip, the CEO there, was walking us through it. We saw Starcloud 1 on the screen, Starcloud 2 being developed, and the early iterations of Starcloud 3, which will go on Starship. We were talking about what actually makes it possible to switch over to compute in space, in orbit. It’s really the economics of Starship, and what that unlocks: an entire market that’s constrained by compute on Earth. Whether it’s what they’re doing with Terafab, Colossus, or all these other projects, I’m curious how you’re thinking about that capacity and the hyperscaler side of things, and when we’ll start seeing compute in space. ### Compute Moves Into Orbit Christian Garrett They’ve publicly shared their timeline: they’re trying to launch some of the first orbital compute satellites in the back half of next year, I believe. They seem to be on track. I think this is another project that’s going to be successful. You can never accurately forecast timelines or capacity, but they have a history of underwriting complex engineering projects at SpaceX, and when they have conviction about something, they deliver it. I think this will be no different. You’ve seen this already. Google just tested a version deployed in orbit recently, and companies like Starcloud are being built out. This is clearly going to be a category and a business that are quite successful. You see all the constraints on Earth, particularly around power, and the clear solution is data centers in orbit. It’s another example of the benefits of SpaceX’s vertical integration. As you said, driving down the cost of infrastructure allows people to build things on top of it. We saw that with Falcon 9: the exponential increase in the number of launches and how much capacity was delivered to space as SpaceX drove costs down. Companies we’ve been invested in for a long time, like Varda, which our friend Delian started along with Will Bruey, have built a successful business on top of lower launch costs. That allows them to deliver valuable capabilities and services to the pharmaceutical and defense industries. I think Starship will drive launch costs down by an order of magnitude, which will probably lead to an equal rise in the number of companies that can be built. It’s very hard to forecast these things, because a lot of the business models we’re talking about today within SpaceX weren’t even thoughts one, two, three, four, or five years ago. That’s the power of vertical integration and the importance of what they’re doing: driving down costs around critical infrastructure. We’re seeing that in AI, but it’s also clearly what they’ve been doing for decades around launch. And it’s still early days as Starship continues to develop. The iteration of Starship we see today is a fully functioning launch platform compared with everyone else. I think the goal is to achieve full reusability, which will have a real impact on cost and launch cadence. But if you compare it with every other launch vehicle today, it’s completely viable. It can launch and deliver payload. The fact that the entire thing isn’t fully reusable doesn’t necessarily miss the benchmark compared with the rest of the industry, because no one else is fully reusable. I think this is probably underpriced. The next phase is getting this thing fully reusable and continuing to open up launch capacity for the market. That takes time. We’re also seeing the early iterations of Starship vehicles. They’re going to continue to iterate, make them larger, and make them more performant. What we see today of the Starship version is definitely not what it will look like five or 10 years from now, in capabilities, size, and everything else. We remain very excited. We’ve been excited for almost two decades. We’re not the only ones. Sean Maguire is so excited he put a model out on Twitter, which was awesome. I love Sean. Molly O'Shea I heard it was the best financial model I’ve ever seen. Christian Garrett So great. I love Sean. That shows the level of excitement among people who have been close to this company. I think now, in the public markets, more people can understand and start to learn this business. If you ask me today, it’s still the company we’re most excited about, and I think the most important company in the world. Molly O'Shea Incredible story, and a demonstration of how to navigate markets when they’re completely changing. On the hyperscaler side, in this neo-cloud environment, I want to dig into this, because I know you monitor all of it. You take an economist’s view, but you also understand the financings behind it. Elon recently put out a tweet about the constraints on financing in this market, the debt, and access to it. We saw this with CoreWeave having some issues, among others. How are you seeing this market play out in terms of financing and debt? Christian Garrett You’ve seen a lot of volatility in the public and debt markets more recently. There’s a lot conflated with this, because there’s also a lot of macro volatility and an election cycle. There’s a generic de-grossing in October that’s pretty consistent, and some of that also plays a role. But there was clearly some questioning of people’s ability to underwrite where revenue is coming from to support the spending commitments, or companies’ ability to grow based on their access to capital to match those commitments. It’s ironic, and it just goes to show you that market structure matters and sometimes these things don’t make any sense. In some sense, the beneficiaries of the spending dropped, but then the spenders dropped too. It’s a little unclear how both companies could drop. One thing to pay attention to is that to build out this capacity, speed and payback period are everything, and cost of capital is a huge part of it. A lot of this is funded by the debt markets because that’s the most efficient form of capital to go do it. You’re seeing people question whether folks can build out that capacity, and I think it’s a valid question. I think some folks may have been treated unfairly. There’s been a lot of noise around people’s access to capital. At the same time, Elon is 100% right that certain companies have a huge advantage and a lower cost of capital. They’re also going to take up a lot of that demand at better rates. It’s another reason I think SpaceX is extremely well positioned in this category. There are a lot of dynamics in the market driving all of this, but there is a real question around certain parts of this market. That question has continued to surprise to the upside if you look over the last few years, because we’ve been asking these questions. But it’s a strange market dynamic where, as time passes, the questions are still there and getting even harder to have a view on. That’s what makes investing in this market so exciting and so difficult. You don’t really know what market dynamics will be like in the next three months, let alone the next three years. It’s also very hard to know steady-state market structures: model pricing, the volume mix of open source or closed source, compute pricing, cost of capital, or the ability to build data centers and access power. It’s fascinating to watch. It’s a fun time to invest, but it’s also incredibly difficult. How to price a company is going to be entirely driven by your ability to believe in some out-years of earnings, and that’s really hard right now. It’s hard to believe in the next two years of earnings, let alone the next 20. I think it’s a fun time as an investor and a difficult time as an investor. Molly O'Shea Yeah, I want to push on that more before we get into the portfolio, because you mentioned a couple of these areas you’re working on or looking out for in market dynamics and structure. Molly O'Shea So of those five or six pieces you laid out, what are the most vulnerable parts? And at the end of the day, is pricing just based on momentum and stories? How much of it is really tied to the numbers? Christian Garrett It’s hard to say. There are areas we’ve focused on with higher conviction, and we’ve tried to look for companies that we felt had more durability and defensibility, which is always our focus on underwriting. We’ve obviously invested quite a bit in Cognition. We’ve just co-led Hex’s round, which is a company we’ve been investing in for a long time. Both of those sit at the harness layer and the app layer. We haven’t done a ton there, but those are companies we felt were pretty well positioned, with third-party structural positioning, which is very beneficial. Not every company wants to be reliant on a lab for certain types of products. We’ve definitely seen that in security and coding, where companies can route to different models. We think it’s also true in analytics and a lot of data workflows, which is where BI and Hex sit. We believe those companies have good structural positioning, and it’s easier for us to underwrite some of the durability of the business. The second thing is that we’re just backing growth. All those companies are growing extremely fast. With companies like xAI and SpaceX, you’ve seen the value accrue to the hyperscaler and the compute layer regardless of outcomes. We believe open source is clearly going to make up a large percentage of models and deployment, just like it does in software. Where do you pick your bets? How do you size those bets? It hasn’t been easy. There are certain categories we haven’t invested in as much, and that’s less a view of how well the companies are doing. It’s just been harder for us to think about and grasp the durability of the growth and what market dynamics will look like in the future. In essence, we’ve made bets in a couple of areas, and in a couple of others we’re still spending time getting more conviction. I don’t think I have a better-informed view than someone else. We’re all spending time underwriting this. I will say, though, that an incredible set of companies has been built at layers where we haven’t invested. OpenAI and Anthropic are two of the most impressive businesses built in a generation, with incredible products that we’re all using today. A moment of volatility or questioning what the out years of growth and economics will look like is going to happen in a fast-moving, transforming market with a lot of disruption. You’re going to have all these things shake out. On the other side of the volatility, you can peer into what’s ahead. Just look at right now: I’m sure we all have family members who are using ChatGPT, and using forms of agents, Muse, Dot, and Grok bots. I’m still impressed seeing cousins, friends, and grandparents using these products. That’s a clear sign that they have a lot of runway to make an impact in the world and are still early in their diffusion. Markets over-extrapolate the good and the bad, so you just want to focus on being right on bets and picking businesses that can survive these cycles. Molly O'Shea What growth rates are you looking at, and how has that changed over time? What metrics are you looking at? Christian Garrett It depends on the type of business. There are businesses that aren’t consumer internet scale and can’t grow as fast as something like that. For us, it’s much more about the strength and quality of that growth. A business like SpaceX compounded at around 40% for 20 years. A 40% growth rate in a standalone year is good, but it’s not incredible. What’s really amazing is the length—the duration. Twenty years is insane. That’s the power of compounding. A 9% growth rate, which is roughly the standard average growth rate for the S&P, over 30 years is about 13x. It’s incredible how much duration matters. For us, it’s mostly about the duration of growth. When we’ve made big bets and really concentrated, it’s in companies where we’ve built conviction over time and understand that the business can compound for a long time. In our asset class, we can’t control duration. I don’t know how long I’m going to be invested in a business. When we make an investment, we want to have a view that we want to be in that business for decades because we don’t control it. We’re minority shareholders, so we want to be excited about it 10 or 20 years from now. The businesses we’ve backed, particularly the ones we’ve concentrated in—SpaceX and Anduril—are businesses where the duration of compounding and revenue growth was very clear, and the runway of opportunity is very large. The growth rate within a year is an important data point, and it supports your math, but it’s not the only factor. Molly O'Shea At the stage where 137 invests, you guys are later stage. There are some earlier bets, but I feel like every company now is at least a $2 billion company. For whatever reason, we moved the floor up to $2 billion over the last couple of companies. I guess everyone’s just going to be a $2 billion company. With that in mind, prices just keep getting pushed up. I just had Shardul from Index on, and he thinks prices are going to continue going up. They’re fine paying a 20%, 25% premium. They’re like, “If it’s 25%, we’re good. If it’s 2x, we need to reconsider.” But there is some momentum play of fueling with capital, and you have a handful of portfolio companies that have been on this capital train, able to snowball more and more money. Cognition was recently valued at $48 billion and is rumored to be in another round to get to something crazy like $70 billion, or $60 billion to $70 billion. Ramp is at $60 billion. I’m sure they’re going to get to $100 billion soon, and Anduril is at $90 billion. We’re seeing a lot of growth externally from these valuations being pushed up and the money being fueled into them. How much of that is a concentration play—these are just clear winners in the market—and how much of that is the growth of those businesses? Christian Garrett It’s a funny quote you bring up from Shardul. It reminds me of a quote my godfather, Teddy Forstmann, had. This was in the private equity industry in the ’80s, and he said, “Everyone likes to walk around and say that they don’t pay premium prices. They pay good, fair prices. Meanwhile, every single day they pay and overpay for things.” In your personal life, you’re overpaying for something all the time. The reality is, you’d much rather own a quality asset that’s a 5x and pay a 25% premium to buy it than own a shitty asset and pay a 25% discount to buy it. You’d want to make that trade all day, which I think is true. Today’s market is a mix of both things. There is incredible growth, and functionally, you have the markets backing that growth. These companies have delivered extensively, building incredible businesses that continue to compound. Christian Garrett Ramp is a company we’ve been invested in for years, and it’s a case study in continuing to expand the purview of the business and launch new products. The core thesis from our first investment has completely played out in its ability to consolidate more and more of the financial operations stack in the back office. They’re continuing to move into more things. They’re still under-levered in a lot of parts of the business, meaning they’re underpricing things and not monetizing them as much as they could. There are other products they can move into that they haven’t yet. There’s a lot of runway for this business, and you’ve seen the investor base continue to back that growth. At the same time, a lot of capital is flooding into certain categories, and there aren’t many companies, particularly in the growth stages, to back. That leads to pricing and market dynamics. In order to access these businesses, you’re going to have to pay a certain price, and people are willing to clear it when there’s a lot of capital. Our asset class always goes through that. There will always be more capital than there is supply of opportunities. But the power law within our industry more than pays for these opportunities, and I think we need to revisit the past to learn these things. There was no shortage of capital flooding into fintech just a couple of years ago, during the COVID boom. A ton of companies are in the graveyard right now. Molly O'Shea They’re all dead. Christian Garrett All dead. Except if you look at Stripe and Revolut: 100-plus-billion-dollar opportunities and outcomes. Ramp is on its way to becoming that size of outcome. Going down the list, the fintech category produced massive wins that have performed extremely well and are all doing billions and billions of dollars of profit. That capital wasn’t wasted. There’s a power law, with only a couple of companies. Nubank, among many others, is another example. There’s a power law, and I think that’s how this will play out across many categories, definitely in AI and in aerospace and defense. So yes, things will be mispriced because there’s too much capital chasing too few opportunities. The power law will probably pay for that amount of capital. As an investor and as an LP, you probably want to make sure you’re in the power law of outcomes, because that’s just how our industry works. Even at growth, it’s the same dynamic. Portfolio construction looks very similar within growth, particularly from a returns perspective. Just a handful of companies drive all the returns, and that’s how we’ve built the firm to capitalize on it. The current setup clearly represents how our industry is set up. Molly O'Shea Let’s talk about Anduril. So, is everything computer? Christian Garrett It’s funny. This is something Palmer’s been obsessed with for quite a long time, so it’s really cool to see him actually do something here. It’s part of an overall trend: repatriating more and more capacity that we should not have given away. This is just another example of that. I particularly like seeing things like this. Sometimes you need industrial policy to bring that capacity, and sometimes companies do it themselves because they see a commercial opportunity. Anduril’s a great example of that, and it’s pretty cool. What I really appreciate is that Palmer’s been talking about this for a long time. If there’s one thing you can appreciate about generational founders like that, it’s that they tend to be early and on the ball. The best part is that, in our job, we get to partner with them in small ways as they need capital and resources to deliver on that. Palmer’s done that continually, along with everyone else in the Anduril founding team. Bringing that computer capacity and manufacturing is another example of that. Molly O'Shea But Anduril’s also in your bucket of conglomerates, so speak to that a little bit more— Christian Garrett Yeah. Molly O'Shea —as it relates to them. Christian Garrett Yeah. As I was saying earlier, it’s something I’ve been more and more obsessed with, and in some ways it led to a big crash in the ’80s. There was a reason those companies were very different. I’d like to return to the conglomerate thesis now. I think today’s conglomerates are different in 2 ways. Traditionally, the conglomerates of the past, like I mentioned, were a bunch of disparate subsidiary-type businesses. You had the benefit of exposure to different industries, real scale, cost-of-capital advantages, some operational advantages, and diversification, which I think is valuable. The negative, in some ways the inverse of those benefits, was that because they were so diversified, it was hard to do a sum-of-the-parts analysis and value these things. You didn’t have a lot of operational synergies among these businesses, so there was no shared common operating platform. They were super disparate. Maybe you started to get diseconomies of scale, and certain segments of those categories were significantly less valuable than others, dragging down businesses that were otherwise profitable and doing well. Today’s conglomerates look different. One, they actually have a core common operating tech platform, so there are synergies in their operations. Sometimes they’re vertically integrated. Look at Apple, which has a ton of consumer products and has vertically integrated on top, building a massive business across the App Store and the Apple layer as a platform, with its own services on top. SpaceX is an example of both. It’s vertically integrated, even though it has completely different industry exposures and business lines, all built on the same platform. It also has a common operating layer, and it uses technology to manage disparate businesses, particularly on the AI side. That’s worked very well. Anduril is another version of this. I call it a lowercase-c conglomerate because it’s not across sectors like traditional dual-use conglomerates, which I want to get to. It’s going after what most of the largest primes are: multi-product, multi-capability, multi-system businesses. What has made Anduril unique, and I think given it economies-of-scale benefits even at the operating level, is that it uses technology extensively in its operations. It’s also built the same core backend infrastructure from a software and AI perspective: Lattice, which powers kind of everything. Every marginal product sold makes the value of the other ones go up. That’s been critical to the story of the business and the investment thesis when we were spending time there back in the day. I think it’s what differentiates them. From a resiliency perspective, and in terms of where value will accrue, I think we need to return to dual-use conglomerates. Quite frankly, I think Palantir and SpaceX are the only examples of this. General Mills made cereal and weapon systems. That’s something I think is important. Christian Garrett China’s defense primes get 60 to 80 percent of their revenue from the commercial sector. Shyam Sankar from Palantir talks about this all the time. The majority of Palantir’s revenue came from the commercial sector in the last earnings report we saw. The majority of SpaceX’s revenue comes from the commercial sector. I think we’re going to need these types of businesses, where you can get real scale because of exposure to the commercial sector. I think the markets will value the diversification, and that scale allows you to subsidize and maintain capacity and capabilities that are critical to the defense sector, as well as throttle up and down as needed. Industrial policy can only do so much, and some of what we’re doing is really good. But, quite frankly, to have the type of capacity we really need from a defense perspective, you’re going to need to be dominant in the commercial industry, or else it just won’t work. Shipping is the greatest example of that. China’s leading prime there is so dominant because it’s the dominant commercial ship manufacturer in the world, and it also makes defense products. I think we need to see more of that. I say a lowercase-c conglomerate is the best way to build a defense business, or any type of business in your category, by being multi-product. It just increases the TAM, and you can continue to expand and take in more and more of an opportunity set. Christian Garrett I think, from a larger point of view, we need these capital-C conglomerates that span the commercial, private, and public sectors. They’re selling across commercial markets and government in order to build the scale we need, which I don’t think the government will be able to subsidize alone. Like I said, Palantir and SpaceX represent that. I think they’re next-generation companies that are going to embrace it. A lot of defense companies aren’t building the first, lowercase-c conglomerates. That’s why I think Anduril is going to be by far the power-law outcome in this category. A lot of companies also aren’t building real, true dual-use capabilities. The best success stories are probably companies that are dominant in the commercial sector and then cross over into the defense industry. You see a lot of that in enterprise software, infrastructure, and cybersecurity. I think we’ll need to see more of it in aerospace and defense. Molly O'Shea One of those next-generation companies is Onodrim [?]. We had the pleasure of interviewing Isis [?] at the Strike conference in Monaco. It was technically Cap Frog [?], but he had some great points and was very spicy. He was the first—or second—employee at Palantir in Europe. Christian Garrett He was second, yeah. Molly O'Shea Second employee at Palantir in Europe. You co-founded this company. Can you walk us through the story of Onodrim and the big announcement you’re coming out with? ### Building Europe’s Defense Suppliers Christian Garrett Yeah. Onodrim is an aerospace and defense scale-up platform. We want to use technology to buy and acquire businesses across Europe in critical industries, deploy technology into them, scale them up, and grow them into very large businesses. There are businesses in the US that do this successfully, including supplier consolidators like TransDigm, Teledyne, HEICO, and Honeywell, across different capability sets. There have also been very successful roll-ups in defense that acquire smaller systems and scale them up. And, obviously, the primes in the US are very acquisitive. A lot of the defense industrial base isn’t made up of the large primes. It’s the thousands of suppliers across the stack: smaller complementary systems and other capability sets. There are also dual-use companies that are dominant in commercial markets, but 30% of their capacity or revenue goes to the defense industry, providing pretty critical capabilities. Most of that has either been consolidated or owned by private equity. In Europe, you haven’t seen that level of consolidation. It wasn’t a growth market, for one, so the primes weren’t very acquisitive. Private equity wasn’t very acquisitive either, and there were no roll-ups—or a European HEICO consolidating suppliers and services in these categories. That’s surprising, given that Europe has a mature aerospace supply chain because of Airbus, valuable capabilities for both the European and US defense ecosystems, and an extremely strong industrial sector. That’s the opportunity now, as spending flows into Europe and countries rebuild their industrial bases. The rest of the defense industrial base consists of hundreds of thousands of small companies and businesses that need to be scaled up. The challenge in Europe is that it’s fragmented, not a single market, which makes it very hard for these businesses to scale. It’s also hard for them to adopt technology or find partners with the right incentives to help them scale their businesses or expand their products. We partner with and acquire these businesses, and we’re long-term holders. We partner with management and the teams, and we invest in them—whether that’s capital expenditures, new products that we can support with engineering resources and capital, or technology to modernize their manufacturing and operations. We help them scale into Pan-European, and hopefully global, champions. We think supporting reindustrialization and rearmament in Europe is the most important mission right now. I think it’s critical for the West. I’m half European and half American, and I think it’s deeply important for America that this succeeds, as well as deeply important for Europe. It also represents where we think a lot more value is going to accrue. Right now, a lot of money is chasing the neo-prime route, and we think the players and winners are established. Anduril is the largest one, and we benefit from that. Even in Europe, if you look at Stark, Quantum Systems, Helsing, and Cambridge, there are all these companies that have received backing and have some structural sovereign advantage within their countries, which have large enough budgets. Now there are 4,000 drone companies in Europe. Clearly, this is overfunded and too crowded. What hasn’t received as much support is the layer below: who’s supplying all these companies? Where are they getting their key components, their aerostructure manufacturing, and their engines? All of that is a layer below that needs to be scaled up. We also want to partner with dual-use companies that are large in the commercial sector and maybe repurpose them for defense, or that have most of their revenue in one sector or the other. I think that’s important for a business’s durability, and for maintaining capacity. Industrial policy alone isn’t strong enough to support the economics needed to maintain the level of capacity we need. That’s the platform we want to build. Most of the business is focused on buying suppliers and services, consolidating that market, and scaling it up. We’ll also look at other opportunities, such as smaller complementary systems on the defense side. It’s funny: although our concept is different, there’s been a lot more excitement about this. That’s not just in the AI roll-up world, which is very different and focused on automating services. Amca [?] and Valinor, who are big fans of Jay and Julie, are building great businesses here. They’re solving different parts of the same issue everyone is recognizing: there’s a layer below, a universe of smaller companies that are critical to the defense industry. Funding and talent have been flooding toward one area, and now too much is chasing it because people saw a big outcome and think they can have one too. I think this layer is underappreciated, and we want to go after it. That’s what we’re doing. We’ll announce some of our first acquisitions soon, and we’re excited to be a really big partner in the ecosystem. We also have partnerships with leading companies, including prime-type companies across the defense and commercial sectors, that we’ll announce soon. Hopefully, we can build what we think will be Europe’s largest technology scale-up platform focused on this sector and support a big mission. Molly O'Shea How do you access talent in this underappreciated layer? Christian Garrett There are two layers of talent for us. There’s talent we recruit into the company, which we’ve done a great job of and will continue to do. If any talented Europeans want to work on this, please reach out. We’ve brought together some very unique DNA. Our company consists of people from private equity, traditional aerospace and defense, government, and the technology sector. We have people from Palantir, SpaceX, Rheinmetall, McKinsey, and others. I think that’s really unique, and it’s been fun. We also have people from nine nationalities, which I think deeply represents Europe. That’s been fun too. It probably makes our Slack and group chats very interesting, because we have a lot of views and cultural differences that we learn from and appreciate, while bringing everyone together. Molly O'Shea Some people are a little more direct than others. Christian Garrett Yeah, definitely. I’ve learned a lot about Eastern European humor. Molly O'Shea Oh, really? Christian Garrett And I think a lot of people have learned a lot about Italian schizophrenia. So it’s been good. What matters most are the companies we partner with. They’re run by people who have built incredible businesses, many of them multigenerational or around for over a decade. We’re looking for management teams and companies with the characteristics I look for as an investor. First, these are durable businesses. What’s the quality of the revenue? Have they been doing this for a long time? Is churn low? That shows it’s a high-quality business. Do they have capacity that’s critical to the industries we care about, particularly on the defense side? Even if most revenue comes from aerospace or industrials, is there a business line that we think should and can grow on the defense side? Does it solve a critical bottleneck? That’s something we really focus on: where the bottlenecks are for this build-out. Contracts have been going to the primes, but production is the big bottleneck right now, partly because of suppliers. Can they scale? Who will take the risk on CapEx? How do we modernize operations to deliver faster? A lot of capacity has been repatriated away and no longer exists in Europe, and you see the same dynamic in the US. We also want to partner with management that knows its category extremely well, and empower them. Many of these companies have those characteristics. It’s been great working on this business over the last year, and we’re excited to partner with the first set of businesses. We’re very early in proving it out. We have a strong thesis, an incredible team, and great backers. Now we need to execute. Molly O'Shea Are you having fun? Did you ever think you’d co-found a company? Christian Garrett Not at all. Molly O'Shea Do you think this has helped you become an investor or a better investor? How do you think about all this for yourself? Christian Garrett It follows a similar MO to another company that my partners helped co-found, Alex and Justin, and helped put together and incubate: Armada, which Dan Wright and John Runyon founded. You know them really well. That initially started with my partners Alex and Justin coming up with an idea they wanted SpaceX to do. SpaceX said it was viable and a good idea, but they had too much going on, so we should just go do it. Obviously, we don’t start companies. That’s not really our MO. That led to conversations with Dan, who was thinking about something related. Dan spoke to John, spoke with SpaceX, and then Dan really ran off and started the company with John. It’s done extremely well. This is somewhat similar. I was spending a lot more time in Europe, working with a lot of our companies there. I had some good relationships out there, and that led to learning the ecosystem. It became pretty clear that this was one of the biggest problems there and was critical to being successful. It wasn’t going to be about how much money Europe spends. It wasn’t going to be solved the way I think it was: “Hey, let’s just give a bunch of contracts all across disparate systems. Let’s have multiple different types of systems, and production’s going to be very slow.” You have a window right now where you’re just going to be behind and very vulnerable. So what’s the best way to support that from a mission perspective? Clearly, it was solving this production issue. That led to going down the rabbit hole more and more. The unfortunate reality was that I may be the person to put this together, probably because I was close to the problem and the ecosystem on both sides. The defense ecosystem consists of the U.S. and European prime and neo-prime companies. Those were all the products being bought, and I was close to that ecosystem. I was part European as well. That led to talking to Isis and Alex, who were extremely well-known, well-respected, and well-regarded in our circles from the Palantir side, given we’ve been invested in that company almost 20 years. That led to the formation of Monodrom. In essence, no, I did not want to start a company. It’s not what I think is my superpower skill set. I’m glad to be a minority co-founder in the company. Isis and Alex are building it, and I help where I can. If there’s a great idea, most times you’re not the person to go do it. If it’s a weird case where you’re the person to go do it, then by all means, go and do it. I’m open to that. I treat it like I’m the most helpful and involved investor, but my workflow’s the same: I just help where I can. I try to get deeply involved. I will spend a lot of time, but I don’t rest on my laurels as an incredible operator, or want to go run these things. That’s not my MO. I love my skill sets and superpowers as an investor, and I think I can apply that as a co-founder in a company if it seems that I need to be. It’s been exciting to go work on something that I think is important and work closely with the ecosystem, which we get to do. The primes and these large commercial companies that I have close relationships with, that we have close relationships with, are our customers most of the time based on the companies we’re buying, or our partners sometimes on opportunities. That’s been pretty fun. The most valuable thing, as you said, which I didn’t appreciate—unlike my friends, like Trey, who’s a mentor of mine and has always talked about this, DeLeon, who’s one of my close friends, and Scott Nolan, the first person I met in venture. These are people I’m extremely close with. Josh has co-founded a company from Thrive, and you can go down the list. A lot of people have, in our world as partners. Molly O'Shea Did you have FOMO? Christian Garrett I didn’t have FOMO. What I appreciated more was that I knew they viewed the opportunity as the most intellectually fulfilling thing, because you got to go really deep. In venture, you get more breadth and pockets of depth, but you don’t get to go multiple layers deep. Trey talks a lot about that. It makes you a better investor, 100%. There are things you understand, appreciate, and know that definitely make you better at the job. I think you can also relate to founders in a different way. There are different layers of pressure, right? Not being the CEO of a company is different from being the CEO of a company, obviously, but I think there’s a lot to relate to. It definitely has. I think the biggest thing is it’s made me appreciate the annoying things that venture investors do. Molly O'Shea Really? What do they do? Christian Garrett Number one, we get a little manic sometimes and worried. There are just things like that. That’s probably the biggest, but there are a lot of things I’m really glad I’ve been able to see from the other side. Molly O'Shea Do you find yourself withholding from certain actions you used to do beforehand? Christian Garrett Definitely. Molly O'Shea Yeah? Christian Garrett Some of that’s with age, for sure. You just learn. Molly O'Shea You’re in your 50s now, is that what it is? Christian Garrett No, but you start in your 20s and you’re trying to make it happen. You’re younger. I think there are a couple things I appreciate more, for sure, and there are also things I understand now on both sides. When fundraising, I understand how investors are thinking and how to talk about the company and all that. It’s great because it’s just a very honest and real conversation. People who’ve done this before me have known this for a long time. When talking with founders about companies, you understand what’s actually going on, what the real story is, and what’s difficult. You understand what’s most helpful. At the end of the day, venture investors can be helpful at times. We try to be, as a platform. We embed pretty deep into our companies. Christian Garrett We have operating partner Ashley, who goes and works at our companies. We do a lot, but we know that at the end of the day, that’s not our MO. We back people who know what they’re doing, and that’s it. A bad investor can destroy your company. A good investor won’t make your company, as my partner Justin always says. At the same time, there are areas where you really want your investors to be helpful, and they’re simple. It turns out you need revenue and talent. It’s, “Can you help with these?” A lot of investors promise they’ll help with things, and they really don’t. There are various reasons, and none of that is nefarious. I think you learn these things more deeply when you’re on the other side. It makes you a better founder and a better investor. For me, as an investor, I’ve been grateful to get some operating experience and exposure, and hopefully to support Isis and Alex in the company and help us deliver for the European ecosystem. I think about the defense industrial base of the West, because a lot of our partners are U.S. companies we sell to and partner with. It’s a big mission, and we’re excited to do it. It’s also affected my thesis in other areas. It’s made me more excited about things being built here. As I mentioned, AMCA and Valinor are really exciting companies. It’s made me more excited about deploying technology in industrial manufacturing. We’ve been close to it because Palantir really pioneered deploying software into legacy manufacturing, from shampoos to planes. Companies like Hadrian and others have started adopting these things, and you learn there’s low-hanging fruit, like automating paperwork. But there are also incredible things you can do now, particularly as AI improves. You can integrate into PLCs significantly faster, which lets you do a lot of interesting things with machines. It’s made me much more bullish on how technology can be used in manufacturing and industrials. I could go on about other things, too. Molly O'Shea Sure. Christian Garrett That’s kind of helped with some of the conversations we have as investors now at 137 Ventures. Molly O'Shea Yeah, I know we didn’t get to talk about them, but Applied Intuition is also an incredible example of this. Christian Garrett 100%. Molly O'Shea …example of this. Christian Garrett Yeah, 100%. They’re a company we’ve been investing in, and we’re really excited about them. They’re solving hard problems and deploying technology into a legacy industry, but one of the most massive industries in the world: automotive. They’ve expanded across industries, across all things robots on wheels and vehicles, and across the OEMs. Construction, mining, trucking—that’s a business as well. We were just at their offices the other day, and it’s incredible to see a company not just deploying and selling software to engineers across these companies, particularly in automotive, but also touching the electronics of these vehicles. They’re simplifying all the processing units, controllers, and wiring, and improving the cost structure of these vehicles. That lets you use more technology in the vehicles and have a better experience, whether for maintenance, your in-cabin experience, or deploying autonomy. It’s been incredible to see people not just take the operating playbook of venture, but also find technology in hard industries. It continues to surprise me how far that can go. You might think the automotive industry is one that’s been successful, but if you look at General Matter, another company we’re really excited about and have been investing in, I mentioned that Scott was the first person I met in venture, a long time ago. I could go on, but it’s a similar thing: they’re bringing back an important, critical capacity in uranium enrichment, while reinventing the playbook for how to do it and applying technology in ways you wouldn’t expect. Even just applying sensors to things that didn’t have sensors. Most of this capacity is with state-owned actors still using Cold War technology. Now we can rebuild some of it and apply the latest tooling. It turns out we can reinvent uranium enrichment, too. It makes you appreciate something Scott said, which is really good: you realize how incredibly smart people were back in the ’60s. Molly O'Shea I think about this all the time— Christian Garrett It’s insane. Molly O'Shea —and it’s so simple. Christian Garrett They did a lot of this— Molly O'Shea I woke up this morning and couldn’t sleep, and I was thinking about how violins were made. I was like, “Why am I thinking about this?” Christian Garrett These people were definitely smarter than us. They— Molly O'Shea But they made violins without us, and now I’m thinking, “Oh, well, AI can generate new types of instruments.” Christian Garrett They were definitely smarter than us. They were doing things by hand and with instrumentation. It’s incredible. But we have new types of tools and technology that can help us continue to drive performance, unlock new capabilities, and make these things more efficient. That’s the trade-off. It is pretty incredible that so many of these things work and were done without the modern tools we have today. ### Technology Rebuilds Industry That’s another example of a company using technology to improve one of the hardest types of processes in the world. There’s no limit to this. How that plays out in margins, business quality, and outcomes varies. Just because you deploy software into a business doesn’t necessarily make it a technology business, but it has real impact. As we’re reindustrializing, this is the fun part: we get to invest in and partner with operators who get to build. We can reinvent how we did things in the old world. I think that’s the exciting and difficult part. Molly O'Shea Yeah, it’s so cool. I know we’re a bit over time. It’s 9:06 right now, so we’ll try to wrap up in about five minutes. There were a couple of other things on your list that I think are quite interesting. I’d love to quickly go over the trend of political weaponization, and also how lucky we are in tech not to be zero-sum. Christian Garrett I’ve been thinking a lot more about these things, mainly because they’re things I don’t want to see happen. They alter what was really special in the industry and what allowed it to continue growing, gain more and more share of the global economy, and touch, impact, and transform many types of industries. Technology now touches all sectors: energy, manufacturing, and defense. I’ll start in reverse, but I think our industry is unique in that it’s not zero-sum. It’s something we all know but don’t think about a lot. If you go into other industries, you realize how different the culture, behavior, incentives, and ways of doing business are. It’s wild. It’s a small difference, with nuance, but I think it really comes down to being zero-sum. The pessimistic take is probably that the only reason we’re not zero-sum is because we’re in a high-growth industry, and many of these industries aren’t. But I do think it’s a little bit cultural, too, and that’s deeply important to preserve. The more I see petty fighting and arguing within our ecosystem, the sadder it makes me. Maybe I’m naive. I came into the industry as a fan from when I was a kid, looking up to some of these people, and I just don’t like it. I see what happens when an industry stops growing and loses its zero-sum mentality, and it concerns me. I don’t think we’re there, but it’s something we should cherish and protect. The other side is political weaponization. You started to see this with social media. That was the beginning of more concern about how we interact with these companies that have such critical power in this category. I think it’s starting to turn, where now it’s becoming politicized: if you as a founder or a company are on one side or the other politically, then we need to do something about it. I think that is really disappointing. You’ve seen it on both sides, and with companies we’ve talked about today. One reason this is incredibly misguided is that it’s ironic: these companies are usually made up of people, management, and founders with completely different political views. They’re working on nonpartisan things that benefit the entire world economy. They’re some of the largest job providers in the world, and they’re pretty critical to national security and resilience. If you didn’t have some of these companies, we wouldn’t be competitive at all. We’d be totally screwed from a GDP growth perspective, from a jobs perspective, and from a capabilities perspective. I think it’s a bad precedent to set. You’ve seen it evolve slowly, whether it was “We need to break up big tech” not too long ago and regulate them for social media, people burning Tesla cars down, which is insane, or people flipping out at Anthropic and saying, “I won’t use their product because of this.” There’s a weird nuance where these worlds are blending more, so maybe that does mean we need to rethink this. Maybe I’m wrong. But I do think it’s a dangerous precedent, too. Consumer and policymaker decisions shouldn’t be influenced by the politics of the people involved in these companies. They should be influenced by the merits of the business, its importance, and the product it offers to the markets. The markets have their own mechanism for sorting these things out. If people don’t like a product, they don’t have to buy it. They don’t need to burn the product down or try to kill the CEO. That’s a little much, but you can just not buy the product. There’s a good feedback loop. You can choose not to use a model. You don’t have to use Facebook, it turns out. No one’s forcing you to. That’s my view. I think it’s open for debate, and there are nuances. I’m not saying I’m 100% right. But, generally, we really do not want a weaponized political attack on companies in our sector, and we don’t want to support it on either side. We’ve benefited from not having a zero-sum mentality, and we never want to lose that. A lot of that is because the industry has been growing, but I think a lot of it is cultural. Those are awesome things. At the end of the day, we’re all in the same industry, driving real value for the world. Tearing each other apart from within is petty and stupid. I’ve seen it in many sectors and companies. And maybe this is a warning to the outside: tearing this industry apart for political reasons will backfire. I’d like to see the tone calm down a bit, because right now is one of the most important time periods in the world. We literally had a flying car just a couple of weeks ago, and yet we also have all this volatility. I’d like to see things change, but maybe they won’t. For me, those are pretty critical things for our industry to cherish. Molly O'Shea Quite the thought-provoking way to end the conversation, which has been awesome. Really thoughtful. I appreciate everything you shared, from the portfolio to SpaceX to how you think about conglomerates. All of these things are really interesting. I do think we just live in a terror society, and we live in this world where we have too much access to information. We were talking about this beforehand. We don’t need to know all these things. We don’t need to be involved in all these things. How about we just had one newspaper that gave us the news each week? That would be a little more peaceful, but we don’t have that. Anyways, Christian, thank you so much for taking the time. I love catching up and seeing what’s going on in the growth-stage markets with you, and I’m excited for the next Hill & Valley. When’s that? Christian Garrett Spring. So March or April, usually. It depends on the congressional calendar, but we’ll have a date soon, usually by the end of the year or early next year. It’s been awesome. We definitely need to have you there next time. We’ll do a lot together. Delian and Jacob and I are super excited, and we have some good things in the works that we’re excited to announce soon. Molly O'Shea Cool. Awesome. Well, thank you, Christian. Christian Garrett Thanks, Molly.