- World Fund’s AGM and Summit 2026 – themed ‘Built to Scale’ – took place on 9 June at the Delphi Filmpalast am Zoo in Berlin, coinciding with SuperVentures Berlin week, Europe’s most concentrated gathering of early-stage VC.
- The closing panel, moderated by TFN’s Akansha Dimri, brought together Jean Latapy of Lithosquare and Daniel Rothmund of Hyperscale Power to debate what Europe needs to produce sovereign industrial tech champions.
- The room was asked whether Europe would produce a $100 billion industrial tech company in the next decade. Only 10% raised their hands, while OpenAI is aiming for a $1 trillion valuation and confidentially filed an IPO, the same morning.
The day began with Levin Holle, Director General at the German Federal Chancellery, and ended with a three-minute film designed to change the way you see things. In between, the World Fund AGM & Summit 2026 – held at the Delphi Filmpalast am Zoo in Berlin on 9 June, during SuperVentures week – put the question of European industrial sovereignty directly to the founders who are trying to answer it.
The afternoon closed with a 35-minute panel titled “From Breakthrough to Sovereignty,” moderated by Akansha Dimri, founder and Editor-in-Chief of Tech Funding News. The two panellists were Jean Latapy, CPO and founding team member of Lithosquare, and Daniel Rothmund, co-founder of Hyperscale Power. Both companies are bets on infrastructure that underpins the AI and energy transition.
The opening provocation: 10% of the room
Before the introductions, Dimri asked the room to raise their hands if they believed Europe would produce a $100 billion industrial tech company in the next decade. Roughly 10% did. The other 90% did not – or were not sure. The moment was deliberately chosen: OpenAI had filed for an IPO and was reportedly aiming for a $1 trillion valuation, that same morning.
“90% believes the rest,” she said. “So I’m just thinking: what are the reasons that Europe is still confused about whether a $100 billion company is possible, while OpenAI is just aiming for a $1 trillion valuation target?”
The question sits at the heart of a debate TFN has tracked closely. Europe’s deep tech flywheel is gaining momentum – but the gap between scientific leadership and industrial scale remains the defining challenge for the continent’s startup ecosystem.
It was a frame that held for the rest of the conversation: Europe has the talent, the science, and the regulatory architecture. What it has consistently failed to produce is industrial scale. The panel was a ground-level interrogation of why, and what it would take to change it.
The two companies: what they are building and why it matters
Lithosquare is a Paris-based startup, founded in 2024 by mining engineer Aymeric Préveral-Etcheverry and Simon Leclair, that deploys AI and geologist-led intelligence to compress mining exploration timelines. Jean Latapy joined as one of the four founding team members, serving as CPO. Over 1,000 new mineral deposits must be discovered by 2040 to meet the electrification and digital infrastructure demand ahead – but it still takes 7 to 15 years from exploration idea to deposit discovery, with geologists spending approximately 80% of their time on data processing. Lithosquare’s platform cuts that from months to days, with partners already active across the US, Europe, Africa, and Latin America. The company raised $25 million in May 2026 co-led by World Fund and Kindred Capital, with Daphni, Omnes Capital, and Ovni Capital also participating.
Hyperscale Power is building next-generation solid-state transformers – SSTs – that are ten times smaller than existing solutions, with an energy-conversion efficiency of up to 98.5%, far higher than today’s standard. The bottleneck it is solving is physical: the power demand of new GPU models has increased by a factor of 100 in the last seven years, but the transformer technology delivering that power has barely changed since it was invented more than a century ago. Rothmund’s analogy landed with the room: “Having a laptop, and then having a charger that is five times bigger than the laptop. You would say, I don’t want to carry this, it’s too expensive. That’s the same thing we do, but at a big scale.” Hyperscale Power raised €5 million in seed funding in March 2026, co-led by World Fund and Vsquared Ventures.
On the origin stories: From academy to company
Both founders were explicit about the gap between European research quality and European commercialisation. Rothmund completed his PhD at ETH Zurich, where he built a solid-state transformer with 99.1% efficiency, before working at ABB as a scientist and principal scientist. He described the frustration of building something at academia that “goes to a drawer.” The motivation, he said, was simply to get it done.
Latapy described joining Lithosquare as the fourth hire, when the team was four people in an office trying to figure out what they were building. What drew him in was not the technology pitch, it was the ambition. “The idea was not to build a new GPT wrapper, not to build a new pattern recognition system, but a holistic system built on geological science, built with geologists.” That distinction, building a system with domain experts rather than for them, recurred throughout the conversation as a defining characteristic of deep tech done right.
On scaling without losing what made you
Latapy offered the most practically useful framework of the afternoon when asked how Lithosquare navigates the tension between constant innovation and scaling product.
Three things, he said. First, always build the first step with step ten in mind, never let early architectural decisions become blockers later. Lithosquare rebuilt its entire document feature extraction pipeline despite having millions of data points already processed, because the old approach would not scale. “We had to restart on a solid page. It’s a hard thing to do, but sometimes it enables you to move fast on the long term.”
Second, maintain the hiring bar at employee 50 as rigorously as at employee one. “Otherwise we would be diluting the essence of the company.” Third, build a culture where you are willing to break things that are working, including a willingness to make 180-degree product shifts when the evidence demands it.
On investors: European first, then US inevitably
Rothmund was candid about the capital reality for hardware deep tech at scale. European VCs are the right early partners, they understand the regulatory environment, the customer landscape, and the long development timelines that hardware demands. Hyperscale Power’s hyperscale customers – the large data centre operators – are predominantly US-headquartered, which will eventually pull the company toward US capital. But Rothmund was direct about the threshold: “When we talk about a few years in the future, you’ll need capital beyond $100 million just to ramp up on capacity and production. Then there’s almost no way around going to the US, unless maybe governmental funding and grants.”
Latapy described a deliberately European founding strategy – Paris-based, French VCs early, now backed by World Fund and Kindred Capital – precisely because Lithosquare’s mission of reducing dependency on Chinese rare earth extraction is itself a European strategic interest. The investors, he said, share the vision of a global company with European roots.
On what Europe must change: one market
Asked to name the single thing that would most improve the European deep tech ecosystem, Latapy gave the clearest answer of the afternoon: one market that actually behaves like one market.
“If you think of a company settling in the US, one market – directly entering an enormous market. For European startups, one by one, having to put into account the difference in policies, in rules, even very basic stuff in how to settle down. It would typically be the unification of the geological services in each country in the European Union, which have different rules, different processes, different ways of managing their data. Basically, we are exploring one country at a time, when we could be exploring Europe.”
Rothmund pointed to capital availability at scale – particularly the absence of European investors willing or able to write $100 million-plus cheques for hardware production ramp-ups without routing through US institutions.
The advice: Rip the T-shirt apart
The panel closed with a question directed at founders in the room who are three years behind where Latapy and Rothmund are now. Both answers were unusually direct.
Rothmund described being in industry three years ago, ready to move fast, with people “pulling my T-shirt, holding me back.” His advice: “Find a way – either rip the T-shirt apart and go for it.”
Latapy’s was more considered: pick something that matters deeply to you personally – not for the market size, not for the investor pitch, but because on the hard days when things start to break, you need a reason that predates any slide deck. “It’s a great way to keep the focus on the bad days when things are trying to start to break a bit apart.”
The broader summit context
The panel was the final session of a full-day programme that included a morning keynote from Levin Holle of the German Federal Chancellery, an afternoon keynote from former German Vice-Chancellor Robert Habeck, and a keynote from Philippe Tibi of the French Ministry of Economy. The summit also featured a World Fund portfolio showcase and an AI & Geopolitics panel.
World Fund closed the day with a provocation of its own. Daria Saharova, Managing Partner at World Fund, suggested rebranding venture capital altogether: not “risk capital,” not even “venture capital,” but “power capital.” The framing was deliberate. The companies in the room are not bets on software features. They are bets on whether Europe produces the infrastructure the next century runs on – or imports it.
This article is part of our editorial partnership with World Fund.
