Teahose.
SIGN IN
NEW HERE — WHAT TEAHOSE DOES
We read the entire AI & tech firehose — so you don't have to.
PODPodcastsAll-In, No Priors, Acquired…
NEWNewslettersStratechery, Newcomer…
PAPPapersPhysical AI research
PHProduct Huntdaily launches
VCInvestor ScoutSequoia, a16z, Benchmark…
CLAUDE DISTILLS →
7 reads, 30 sec each — free, 6 AM ET.
+ a live graph of the companies, people & themes underneath.
HOME/SOURCERY/The $500B Opportunity in Europe’…
POD
// EPISODE
SOURCERY

The $500B Opportunity in Europe’s Defense Buildout

DATE September 30, 2026SOURCE SOURCERYPARTICIPANTS BRUNO WEJCHERT, GEORGE HAJIMINAS, ISIS, MOLLY O'SHEA
// KEY TAKEAWAYS6 ITEMS
  1. 01Volume Beats Exquisite Technology
  2. 02The White Space Is the Manufacturing Base, Not Another Neoprime
  3. 03Cost Curves Must Collapse for Autonomous Mass Systems
  4. 04Two Company Archetypes: Tech/Integrators and Industrial Component Makers
  5. 05Fragmented European Procurement and Too Many National Champions
  6. 06Government Is the Buyer, and Government Money Pulls in Venture Capital
In this episode

1. Key Themes

Volume Beats Exquisite Technology

The central lesson drawn from Ukraine is that mass production matters more than technological perfection. Isis, who led Palantir's commercial business in Europe and now co-founded Onodrim Industries, argues that if forced to choose between technology and volume, volume wins. The pre-2022 European defense industry was built around bespoke systems that cannot be mass produced.

Isis: "You'd rather have a million of something than one of anything" 00:00:16. She adds: "A lot of defense industry pre-full scale invasion of Ukraine was exquisite technologies... dreams of generals that cannot be mass produced" 00:03:45. Her summary of the new reality: "Anything you do has to be produced in tens of thousands, hundreds of thousands or millions" 00:03:45.

The White Space Is the Manufacturing Base, Not Another Neoprime

Bruno Wejchert, who has mapped the defense venture market, sees capital crowding into neoprimes like Helsing while the supply chain beneath them remains fragmented and outdated. The next-generation primes will need suppliers who can deliver at scale inside Europe.

Bruno: "One of the biggest gaps that we've noticed is... a lot of those systems will need to more importantly be produced in Europe. And in order to do that you're going to have to acquire a very fragmented manufacturing base" 00:05:04. He continues: "the white space is very much in the sort of manufacturing gap for supplying some of those next generation primes that are building systems on sort of outdated supply chains" 00:05:34. Isis's Onodrim is the direct expression of this thesis: "we actually look to find the very good manufacturing and critical mission system businesses in Europe, acquire them" 00:01:21.

Cost Curves Must Collapse for Autonomous Mass Systems

Autonomy, robotics, and AI change supply chain economics. Components once priced for exquisite platforms must fall dramatically in cost to make hundreds of thousands of units feasible.

Isis: "The key thing is actually that everything has to become cheaper. It's like if the radio could cost $30,000 four years ago for military, and that was acceptable. This is no longer acceptable when you're talking about autonomous mass technologies" 00:07:30. She also flags cameras as another component category facing the same pressure 00:07:30. George Hajiminas, Chief Strategy Officer at Theon, agrees: "Scale is the most important parameter here. But also having the underlying supply chain networks that allow for this scale to be reached" 00:08:11.

Two Company Archetypes: Tech/Integrators and Industrial Component Makers

Rather than a single hybrid model, Isis predicts a division of labor. Software/AI companies integrate, assemble, and test, while industrial specialists supply components, collaborating in a non-zero-sum way. George sees a synthesis within single firms, as Theon's own evolution shows.

Isis: "there will be two types of [companies]... who focus on the tech, on the software, on AI, on weapon final assembly and make testing... and then there are industrial companies who are extremely good at producing components that those companies making technology need. And the synergy between the two and not a competitive nature but a non-zero-sum nature" 00:11:02. George: "Prototypes don't fight wars. No matter how good they are, they cannot win wars" 00:10:27.

Fragmented European Procurement and Too Many National Champions

Procurement in Europe is not a mirror of the US; it runs country by country or at the EU level, and national protectionism sustains companies that might not survive in a competitive market.

Bruno: "One of the biggest misconceptions is that processes in US and in Europe for procurement of defense systems are similar. They actually aren't. They're quite fragmented" 00:13:40. George: "we have more companies than are needed in a lot of ways, because so many are national champions that will not survive in otherwise competitive landscapes. Whereas in the US all of the US primes seem to have found their niche" 00:13:16.

Government Is the Buyer, and Government Money Pulls in Venture Capital

Unlike a market where private military corporations drive demand, European defense demand comes from sovereign budgets, and the scale of those budgets is what draws venture funding.

Isis: "Government is the buyer. There are these theories that the private military corporations will be the next way to do defense, but that's not yet the case" 00:06:46. She continues: "When there is like a, you know, trillion of dollars being poured into defense, and by 2030, this triggers venture private funding" 00:06:46.

Talent Is a Tailwind, Not a Constraint

Europe's rearmament ambition is acting as a magnet, including for diaspora talent.

Isis: "hiring into defense has been great. Like people like the Europe's ambition to rearm and re-industrialize" 00:05:58. Her own CTO "spent 17 years in the US and now he came back to Germany two weeks ago with his family to strengthen European defense" 00:05:58. Companies with "big ambitions" like Helsing "have no problem of attracting the best talent" 00:06:27.

Europe's Gap vs. the US Is Scale-Up Depth, Not Technology

George pushes back on the notion that Europe is technologically inferior, locating the real difference in the absence of Anduril and Palantir-scale breakout companies.

George: "European defense has to a very large extent similar technological capabilities to the US. I don't place them as far apart as a lot of people might think" 00:12:15. "One main differentiator is the prevalence of startups in the US that have become scale-ups... Anduril being one of them, and Palantir being another" 00:12:15.

2. Contrarian Perspectives

Scale-Ups Over Startups in Defense

George openly says this is contrarian to the VC world. Defense does not need moonshot revolution; it needs proven tech that can scale.

George: "Defense does need more this ability to scale rather than crazy innovation or revolution so to speak on the technology side and this is a very interesting thing to tell to people that are in the VC world... it's not about finding something that's going to revolutionize everything but finding the right fit that has the capacity to scale... So I would say scale-ups over startups in defense is a somewhat controversial take" 00:18:15.

Superior Products From Small Startups Will Still Lose

Conventional venture wisdom says the better product wins. Bruno argues that customers with sovereign risk care more about balance sheets and delivery capacity than product quality.

Bruno: "a lot of companies have been funded with minimal amounts of capital like five or ten million... The problem is if you don't have hundreds of millions of dollars on your balance sheet and the end customer is the government, they have very little belief in your ability to actually deliver and manufacture the problem at scale" 00:14:44. He concludes: "a lot of those companies in Europe are not going to do well even though their product might be deemed superior compared to sort of traditional OEMs" 00:15:14. Meanwhile some players are "de facto king-made like Helsing in Germany" 00:15:14.

Volume Matters More Than Technology

In an industry that has long prized technological edge, Isis argues for ranking volume first: "if you had to pick between the two, you would go for volume. Like volume wins, like large scale" 00:03:19. She also flags that advantage should ideally be "not easily copy-pastable" 00:04:14, a hedge that acknowledges volume alone is a commodity strategy.

Tech Companies Should Not Try to Become Manufacturing Giants

Against the vertical-integration instinct popular in neoprime narratives, Isis says software/AI defense firms should avoid the ambition to manufacture everything.

Isis: "technology companies who do AI and software for defense and connected battlefield have to not want to also be the manufacturing giants of the world" 00:11:02. She acknowledges it is "slightly maybe controversial" 00:11:02.

A $500B European Defense Company Founded in the Last Five Years

Bruno makes a bold, specific prediction.

Bruno: "if Europe can truly do exceptionally well on both manufacturing locally but also in building the intelligence layer for defense, there could be a half a trillion dollar company come out of Europe in the defense space that was founded probably in the last five years or so... I think could be done in the next decade" 00:18:45.

3. Companies Identified

Onodrim Industries

European defense roll-up co-founded and led by Isis. It acquires strong manufacturing and critical mission system businesses in Europe rather than becoming another prime. Mentioned as the direct embodiment of the manufacturing-gap thesis. Isis: "we're not building like another prime on your prime. We actually look to find the very good manufacturing and critical mission system businesses in Europe acquire them" 00:01:21.

Theon

Established European electro-optics defense company (night vision, augmented reality, head-up displays), IPO'd in 2024. Positioned as a beneficiary of the European defense boom and as an example of an industrial company adding technology. George: "we were mostly an industrial company about four or five years ago. The main product we were making was night vision... But since then we've brand significantly new technologies" 00:09:58. On growth: "both through an IPO in 2024 and having a pretty strong growth trajectory since then" 00:01:45.

Helsing

German defense neoprime, cited as the clearest example of a king-made European winner and as a talent magnet. Bruno: "certain companies have already been de facto king-made like Helsing in Germany" 00:15:14. Isis: "Helsing has seen the same. They have no problem of attracting the best talent" 00:05:58.

Anduril

US defense scale-up. Cited by George as an exemplar of a startup that became a scale-up, something Europe lacks, and by Bruno as a company that partnered with government on joint-venture-style initiatives. George: "the prevalence of startups in the US that have become scale-ups so to speak, and Anduril being one of them" 00:12:46. Bruno: "companies like Anduril or Hadrian have partnered with the government in joint venture initiatives to really change how this relationship between government spending and defense OEMs exists" 00:14:09.

Palantir

US software/AI company with a long path to profitability. Cited by George as a scale-up exemplar and by Isis as the source of her core operating lesson. Isis: "Palantir not being profitable for, I don't know, over 15 years is good, and now being great" 00:17:35.

Hadrian

US advanced manufacturing company for defense, cited alongside Anduril as partnering with government to reshape the government-OEM relationship 00:14:09.

Iceye (transcribed as "Isai")

European SAR (synthetic aperture radar) satellite company, cited by Bruno as de facto king-made at the European level for SAR applications. Bruno: "Isai on the sort of European level for SAR applications. And those companies are just going to keep on eating away at market share from traditional primes" 00:15:14.

Brex, Turing, Public.com, VCX by Fundrise, Deel (sponsors)

Sponsors of the episode, not discussed for excellence in the conversation. Named in reads: Brex (finance platform used by Vercel, OpenAI, Anthropic, Granola, Deepgram), Turing (AI training data and RL environments for NVIDIA, Anthropic, Salesforce, Gemini), Public (AI-generated index assets), VCX by Fundrise (public ticker for private tech), and Deel (global payroll and HR).

4. People Identified

Isis

Former head of Palantir's commercial business for Europe, now co-founder and CEO of Onodrim Industries. Mentioned as the guest with the sharpest strategic frame on volume, cost, and company-type division. Her Palantir-derived lesson: "Focus on being important and valuable and the money will come" 00:17:35. Also: "how am I creating value, why we are becoming more important as a company for the world. And then the money will come" 00:17:35.

George Hajiminas

Chief Strategy Officer at Theon. Offers the operator's view of an industrial company adapting to new technology and a realistic view of Europe's capabilities. George: "It is not so much a revolution I would say, it is more of an evolution as new technologies get gradually adopted into existing platforms and systems" 00:10:27.

Bruno Wejchert

Host with Strike and at Villa Charlotte, who has mapped the defense market. Makes the key analytical claims about the manufacturing gap, procurement fragmentation, and balance-sheet credibility, and the half-trillion-dollar prediction 00:18:45.

Alex (Onodrim co-founder and CTO)

Isis's co-founder who spent 17 years in the US and recently returned to Germany with his family to work on European defense. Cited as evidence of diaspora talent flowing back. Isis: "my co-founder Alex CTO. He spent 17 years in the US and now he came back to Germany two weeks ago with his family to strengthen European defense" 00:05:58.

Alex Karp

Palantir CEO, named as the source of Isis's biggest lesson about focusing on value before profit 00:17:35.

5. Operating Insights

Never Anchor to Next Year's Revenue; Anchor to Importance

Isis describes the discipline she absorbed at Palantir of resisting short-term revenue framing in favor of a question about value creation. Isis: "I always remember when people ask you, what do you sell tomorrow, what's your next year's revenue. I think that's what I remember. How am I creating value?" 00:17:35. This is an operating filter for deciding what to build and which customers to prioritize.

Design for Cost Targets Before Design for Capability

Isis's radio example ($30,000 previously acceptable) implies that in autonomous mass systems the unit cost ceiling should be a primary design constraint, not an afterthought. Isis: "A lot of these components have to go like significantly cheaper in price, significantly higher in scale" 00:07:58. Founders selling into defense should benchmark their BOM against a mass-production price, not legacy platform pricing.

Prototype Performance Is Not Battlefield Performance; Budget for Iteration Cycles

Bruno notes that US systems that tested well in controlled environments underperformed early in Ukraine and needed repeated updates. Bruno: "they needed to go through these iterative cycles where they keep on updating them and making them better... they have done extremely well in testing environments but they just have not managed to perform towards the beginning of the war as well as expected" 00:04:25. Operators should plan field-feedback loops and capital for post-deployment iteration.

Raise Enough Capital to Be Credible to a Sovereign Buyer

Bruno's point that governments discount small balance sheets is directly actionable: fundraising size is a sales tool when the customer is the state. Bruno: "if you don't have hundreds of millions of dollars on your balance sheet and the end customer is the government, they have very little belief in your ability to actually deliver" 00:14:44.

Partner Rather Than Compete Across the Stack

Isis's model of tech integrators and component suppliers working non-zero-sum is an operating posture: pick your layer and form durable supplier relationships rather than vertically integrating 00:11:02.

6. Overlooked Insights

Lead Times in Years Are the Real Bottleneck, and They Explain Why Primes Are Slow

Bruno drops, almost in passing, that component lead times can be "in the years," which is why some primes take four or five years to deliver on contracts. Bruno: "The lead times for some of those components could also be in the years, which is why certain primes take four or five years to deliver on certain contracts" 00:05:34. This reframes the prime's delivery problem as a supply chain problem rather than a software or design problem, which strengthens the case for consolidating and modernizing component manufacturers as an investable wedge, and suggests that anyone who can cut component lead time from years to months holds outsized leverage over the whole industry.

Even the Best Startups May Be Structurally Excluded Because the Customer Buys Balance Sheet, Not Product

Bruno's aside that a superior product from a five-to-ten-million-dollar startup may still lose implies an unusual venture dynamic: in defense, the usual product-led disruption path is blocked by procurement risk aversion, so kingmaking by government (as with Helsing) may be the primary determinant of outcomes. Bruno: "certain companies have already been de facto king-made like Helsing in Germany" 00:15:14. For investors, this means picking winners may depend less on product benchmarking and more on identifying which companies governments are choosing to anoint, and on whether acquirers like Onodrim can roll up credible industrial capacity fast enough to serve them.