Antti Vasara has spent decades at the intersection of science, industry and innovation. As former CEO of VTT, one of Europe's leading research institutions, he has seen hundreds of technologies attempt the difficult journey from laboratory breakthrough to commercial success.

Today, as Venture Partner at Kvanted, he works directly with founders building the next generation of industrial deep tech companies. His perspective is shaped by a simple observation:

Great science creates the opportunity. Industrial adoption determines the winner.

We sat down with Antti to discuss commercialization, capital, AI, and what it really takes to build enduring deep tech companies.

You led VTT for ten years and saw hundreds of technologies move from lab to market, or fail to. What separates the deep tech ventures that break through from those that don't?

The companies that break through treat customer problems as seriously as the science. The ones that fail often stay in love with the technology and assume the market will eventually discover them.

The other differentiator is team composition. A brilliant scientist paired with an experienced industrial operator is a fundamentally stronger company than a team made up entirely of researchers.

The biggest mistake in deep tech is assuming the market will discover you. It won't. You need to understand the customer, the buying process, and the industrial reality you are entering.

"The breakthrough teams treat customer needs as seriously as the science."

As Venture Partner at Kvanted, you work closely with deep tech founders. What does that look like in practice?

A lot of the work is strategic. Deep tech founders often have technology that could serve multiple industries and markets. The temptation is to keep every door open. Success usually comes from saying no to four opportunities so the company can win in one.

We spend a lot of time helping founders translate between science, customers and capital markets. Each audience sees the company differently. The challenge is making sure the story remains coherent across all three.

The same tension appears in decisions around fundraising, product development, customer delivery and hiring. Knowing what not to do is often as important as knowing what to do.

Industrial tech founders often have exceptional technical expertise but are building a company for the first time. What blind spots do you see most often?

First-time founders frequently underestimate how long industrial adoption takes. A pilot can happen relatively quickly. Becoming part of a customer's supply chain and purchasing process usually takes much longer. Funding plans need to account for that gap.

The second blind spot is hiring. Early teams often over-index on scientific talent. As companies scale, they need operations, quality, regulatory expertise, manufacturing and commercial leadership.

The same applies to boards. Investors matter, but experienced industry operators often provide some of the most valuable strategic guidance a founder can receive.

Deep tech has long development cycles and significant capital requirements. How should founders think about balancing scientific ambition with commercial discipline?

The challenge isn't balancing science and commercialization. It's sequencing them correctly. In the early days, scientific ambition is the moat. Founders need enough focus to develop a genuine competitive advantage without getting distracted by short-term opportunities.

But eventually every company needs answers to three questions: Who is the first customer? What is the first product? And how does that product create a defensible position in the market? If those questions cannot be answered by the time a company reaches Series A, it is still doing research rather than building a business.

The science never stops being important. It simply stops being sufficient.

"The science never stops being important. It simply stops being sufficient."

You've shaped innovation policy at both Finnish and European level. Where do you see the most promising industrial opportunities today?

The most promising opportunities sit at the intersection of deep technology and industrial demand. Quantum technology is one example. Europe has built genuine strengths in superconducting hardware, photonics and increasingly the software stack around quantum systems. Sustainable materials and bio-based industrial inputs are another. Regulatory pressure is creating real demand for alternatives that can replace incumbent materials.

Industrial AI is perhaps the broadest opportunity. Not consumer AI, but AI applied to manufacturing, mining, energy, forestry and other industrial systems. Foundation models, simulation and digital twins are becoming powerful tools for improving how physical industries operate.

These are precisely the categories where Kvanted spends most of its time: technologies that create structural advantages for industrial customers rather than incremental improvements.

At the same time, founders should be cautious about businesses whose economics depend entirely on subsidies or political support. Long-term success ultimately requires a customer willing to pay for the value being created.

"The most promising opportunities sit at the intersections where deep tech meets industrial needs."

AI is accelerating research, simulation and product development across industries. What should deep tech founders be doing differently today?

Founders should use AI aggressively across R&D, simulation and engineering workflows. In many domains, the productivity gains are already real. More importantly, they need to understand what AI changes and what it doesn't. AI may accelerate materials discovery, but it doesn't accelerate physical qualification. It may reduce software development costs, but it doesn't eliminate industrial procurement cycles.

The winners will treat AI as something that reshapes competitive advantage, not simply as a tool for reducing costs. The question founders should ask is not how AI improves efficiency, but how it changes the structure of their market and their moat.

Deep tech rewards patience in a way most startup playbooks don't. How should founders think about pacing, milestones and momentum?

Deep tech momentum rarely happens naturally. Founders have to create it. Even highly complex technologies need to become tangible through prototypes, pilots and measurable progress. Investors, customers and future employees all need evidence that the company is moving forward.

The right sequence is usually simple: validate the science, then validate the product, then validate the business. Skipping steps rarely saves time. More often, it makes the next fundraising round significantly harder. Deep tech founders often worry about long timelines. They shouldn't. The timeline is the moat. The challenge is creating enough technical, commercial and operational evidence along the way to bring others with you on the journey.

"The timeline is the moat."

And remember: deep tech is a decade-long game. Pace yourself accordingly.