Building a deeptech company is not for the faint-hearted. It’s part science experiment, part endurance sport, and part leap into the unknown. Over nine episodes of our DeepTech Unleashed podcast, deeptech founders from across Europe who are building everything from AI-powered safety systems to quantum computers and gigafactories, shared what it takes to move from the lab to global scale.

Their journeys differ in industry, timing, and personal background. But beneath the surface, common patterns emerge: clarity of mission, obsession with customers, and a willingness to rewire themselves — and their teams — for the long haul.

Origin Stories & Motivation

A few of these companies began with a blank slate and a spreadsheet. For many founders, the idea was sparked by something visceral.

For Ciarán O’Mara, co-founder of Protex AI, it was a workplace accident that gave urgency to the concept of AI-driven safety monitoring. His background in computer vision gave him the tools, while the mission gave him momentum.

Benoît Lemaignan founded Verkor to tackle what he sees as Europe’s “climate sovereignty” problem. For him, low-carbon battery production isn’t just an industrial opportunity; it’s a strategic imperative for the continent’s future.

Others took less linear paths. Marta Sjögren, after a decade in venture capital, switched from backing others to building her own CO₂-mineralisation startup, Paebbl. She brought an investor’s ability to think at scale, but stepped into a space where much was still unproven.

For Péter Fankhauser, co-founder of ANYbotics, the leap was from robotics research to commercialisation. Years at ETH Zurich taught him how to make robots walk; the real challenge would be making them useful enough for customers to pay for.

Cyrille Kabbara launched Shark Robotics after a career in the French military, motivated by the desire to build rugged, mission-critical robots that could operate where it was too dangerous for people.

Jonas Schneider pivoted from his role as OpenAI's first engineer hire — at the heart of Silicon Valley's AI revolution — to founding Daedalus and spearheading Europe's AI-powered manufacturing renaissance.

Philipp Roesch-Schlanderer founded eGYM with the aim of making fitness training smarter and more connected, blending sports science with technology to create a scalable business model that could outlast market cycles.

Jan Goetz, CEO of IQM Quantum Computers, was driven by the belief that Europe needed a home-grown quantum champion — and that the science he had worked on for years was finally ready to become an industry.

Similarly, Sven Przywarra co-founded LiveEO, convinced that satellite data represents an untapped resource with massive potential to transform our lives. His starting point? A university accelerator team and 200 euros in the bank.

As different as these stories are, each began with a conviction strong enough to sustain years — even decades — of experimentation, setbacks, and iteration.

Scaling & Growth

In deeptech, scaling rarely follows a neat domestic-then-international pattern. For some, global reach is part of the business model from day one. ANYbotics had to ship robots to multiple continents in its earliest deals because clients operated globally. As Fankhauser puts it: “If you work with a large company, they’ll say, ‘I’ll get five robots, but I want them on five different continents.”

Partnerships often act as growth accelerants. Verkor’s alignment with Renault unlocked credibility and capital, compressing what might have been a ten-year ramp into a few. On the other hand, eGYM scaled its connected fitness offering not by going gym-to-gym, but by embedding into the networks of major equipment manufacturers.

Others took a more focused route. Shark Robotics initially built bespoke robots for high-stakes situations, but quickly realised the path to sustainable growth lay in standardising a product line. That shift turned custom engineering jobs into repeatable sales.

While the routes differ, the underlying principle is consistent: scaling deeptech isn’t just about making more of the product; it’s about engineering the entire organisation to deliver, support, and improve it at a pace the market will tolerate.

Technology & Innovation

All of these founders operate at the frontier of their fields, but they share a pragmatic view of technology. The “best” solution isn’t always the newest or most complex; it’s the one that works reliably in the real world.

For instance, Protex AI mixes the latest vision-language models with tried-and-tested computer vision, choosing what works best in the real world — even if it means sacrificing a bit of lab theory. O’Mara explained: “For us, AI is not just about novelty — it’s about reliability in the field. Bad lighting, blocked views… You have to make it work where it matters.”

For ANYbotics, the magic lies in fleet learning: every robot’s experience feeds back into a shared knowledge base, improving the entire fleet. That requires as much innovation in software infrastructure as in mechanical engineering.

Daedalus isn't pursuing mass production — rather, it's leveraging Europe's unique strength in high-complexity manufacturing, using AI-driven, asset-light factories to preserve artisan-level quality at scale. It's a pragmatic bet on where Europe can lead.

Jan Goetz, CEO of quantum computing leader IQM, approaches quantum computing with the same lens. His bet on superconducting qubits isn’t about chasing every promising avenue but rather about picking a path that balances performance with manufacturability. “You can have the best idea, but if you can’t make it work, you have a problem,” explained Jan.

What ties these approaches together is a refusal to build technology for its own sake. Tech innovation here is in the service of deployment, not just discovery.

Sales & Go-To-Market

Selling deeptech is slow, relationship-driven work. For most of these founders, the early sales process looks more like co-development than traditional pitching.

Shark Robotics’ first major product, the Colossus firefighting robot, was built hand-in-hand with the Paris Fire Brigade. That collaboration paid off when the prototype helped save part of Notre Dame Cathedral.

ANYbotics took a different tack: putting early prototypes in customer environments quickly, even if they weren’t perfect, to gather feedback and prove value. In industries unused to such rapid iteration, that approach demanded a lot of trust and a willingness to learn in public.

LiveEO quickly realized that refining satellite data generates more value than simply collecting it. "There's no shortage of raw satellite data," said co-founder Sven Przywarra. "The challenge is making it actionable — you need to aggregate it, harmonize it, and prepare it for analysis."

Understanding the decision-making process is as critical as having a great product.

­Fundraising & Investors

Capital requirements in deeptech are enormous. Verkor’s €2 billion raise, IQM’s €200 million, and Protex AI’s $36 million rounds underline the point. But raising the money is only half the battle: executing at scale is the real test.

Marta Sjögren warns against treating investors as all-knowing oracles, arguing that founders must hold the steering wheel on vision and priorities.

For others, the focus is on alignment. Lemaignan sought backers who not only understood the climate tech opportunity but were committed to the long-term industrial build-out it would require.

In every case, capital was treated not as an end goal but as a resource to be deployed deliberately, often in carefully staged phases to balance speed with control.

Challenges & Resilience

Every founder interviewed could talk about a list of unexpected hurdles. For Protex AI, one Canadian client’s site went dark because snow had covered all the cameras - a reminder that even elegant AI can’t outthink nature.

ANYbotics had to shed its academic reflex to keep refining before releasing. The company learned to shift from research perfectionism to customer-centric delivery, accepting that early versions might not tick every technical box but could still create value.

For Lemaignan, resilience is less about reacting to crises and more about pacing for a decades-long race.

The common thread is adaptability and the ability to reframe setbacks as inputs, not endpoints.

Team & Culture

In deeptech, the team is as much a differentiator as the tech. Founders spoke about culture not as a “nice to have” but as a critical success factor, especially in globally distributed or multidisciplinary organisations.

O’Mara views each new hire as a culture-shaper, deliberately adding people who complement, not clone, the existing team. Jonas Schneider also embodies this philosophy at Daedalus, where Silicon Valley's "move fast" culture merges with German craftsmanship to build a team where diverse strengths complement and amplify one another.

At Paebbl, Marta Sjögren emphasises a culture of “high agency” — giving people autonomy to make decisions and act quickly, even in the face of uncertainty, while staying anchored to the company’s mission to remove CO₂ from the atmosphere.

Shark Robotics runs with a level of operational discipline borrowed from founder Cyrille Kabbara’s military background, while eGYM’s culture is infused with the competitive but team-oriented mindset of sport.

ANYbotics set the tone for internationalism from the start by making English the company language and recruiting globally, ensuring diverse perspectives were built in from the earliest days.

Whatever the style, the consensus is that culture doesn’t “happen”; it’s engineered as intentionally as the technology.

What It Really Takes to Build Deeptech

Across these conversations, a few constants stand out. Start with a mission compelling enough to keep you going when the odds turn against you. Build technology for the messy, real-world environments it will live in, not just the lab. Get close enough to your customers to see their problems firsthand. Choose investors who match your time horizon, not just your valuation. And recognise that in deeptech, there are no shortcuts, only deliberate steps, taken over and over, towards a vision that may take years to fully materialise.

Because Benoît Lemaignan reminds us, even the giants started in someone’s garage.

🎧 Listen to all these conversations on your favourite podcast platform and follow us on LinkedIn to catch every new episode drop.

* Some of the quotes have been edited for clarity.