Domain expertise as a defensible position in the Dutch AI sector

The promise of quick profits and low entry barriers created a wave of new ventures. But as the initial hype fades, a sorting is beginning to emerge. A study of 200 funded AI startups showed that 73% are essentially 'wrappers': companies building a thin layer around existing models like those of OpenAI or Anthropic [2]. Their competitive advantage is paper-thin. The underlying technology is not theirs, and the functionality can often be copied within weeks.

The Defensible Position

As the wrapper economy reaches its limits, another type of AI company is emerging. These are businesses that do not start from technology but from a deeply rooted problem in a specific sector. Their strength is not a clever prompt but years of domain knowledge. They use AI not as an end in itself but as a tool to optimize complex, existing processes. And that's where their defensible position lies, their 'moat' as investors call it: the barrier that keeps competitors at bay.

Take Tibo Energy from Eindhoven. Their defensible position is not the AI model, but the combination of energy market knowledge, industrial process optimization, and regulations. Their algorithm 'Alice' forecasts and manages energy flows from solar panels, batteries, and charging stations almost in real-time. The result is that companies extract more capacity from their existing grid connection, alleviating grid congestion.

Remco Eikhout, CEO of Tibo Energy, sees this daily. "Most companies have no idea how much energy they use or where it goes. They only see the bill. You can't fix what you can't see." [3]

This requires knowledge of energy contracts, peak loads, CO₂ intensity, and grid constraints. You might be able to replicate the model, but acquiring the knowledge of how energy infrastructure works in practice is not something you gain over a weekend. The recent €6 million investment underscores the confidence in this approach [6].

At 10x Team, the defensible position lies in the understanding of the recruitment process itself. Their AI recruiter takes over the first interview round entirely, 24/7, and in more than 70 languages. This reportedly saves recruiters four to six hours a day [4]. But more interesting is what they are trying to solve: bias. Traditional interviews are notoriously subjective. One recruiter focuses on different things than another. 10x claims objective scoring by applying consistent criteria to all candidates. Their system scores on execution capability, stress resilience, and problem-solving thinking, not on likability or 'cultural fit' as often judged in a first interview.

Angelique Schouten, CEO of 10x Team, emphasizes that perspective. "It's not about replacing people. It's about enhancing human potential. That's the future we're heading towards. With AI, not for or by AI." [5]

The fact that 67% of candidates take the interview outside office hours shows that it also solves a practical problem [7]. The defensible position here is not the AI, but the understanding of what a good interview is and how to scale it without losing quality.

The common thread? They are all companies employing AI where human expertise is scarce, expensive, or inconsistent. They automate not the simple, but the complex. And that complexity requires domain knowledge you can't get from an API. That's their protection against competition.

The Next Phase

The question is what happens now. Remy Gieling, founder of AI.nl, predicts consolidation. "In the long run, many small companies will merge into larger ones," he says [1]. That's logical. The market for 'simple' AI tasks becomes saturated. Companies are seeking specialized knowledge tailored to their specific needs. That requires capital and highly educated technical personnel. One-person businesses (67% of current AI companies) often cannot make that leap.

But there is also another movement underway. Companies with real domain knowledge are beginning to strengthen their position. Tibo Energy is expanding into Germany and Belgium. 10x is adding modules for fleet charging and e-boiler control. They are moving from a point solution to a platform. That's the next step: from solving a specific problem to building an ecosystem where you tackle multiple problems within the same sector. The barrier for competitors becomes wider, deeper, and harder to circumvent.

The Dutch Advantage

The strength of the Dutch ecosystem lies in the sober application of technology in areas where it truly adds value. It's a culture where domain expertise and engineering mentality weigh more heavily than hype. The AI wave acts as a filter. The companies that remain are not those with the sleekest interface, but those with the deepest roots in a domain. The future of Dutch AI is not determined by who can 'wrap' the best language model, but by who can solve the most complex, sector-specific problems. The true protection against competition is not technology you buy, but expertise you build over years. And that expertise, combined with the right technology, is what remains when the hype is over.

Max Pinas
Founder Dutchstartup.ai & studio hyra