Domain expertise as a defensible position in the Dutch AI sector

The promise of quick returns and low barriers to entry created a wave of new ventures. But now that the initial hype has passed, a sorting process is beginning to take shape. A study of 200 funded AI startups found that 73% are essentially 'wrappers': companies building a thin shell around existing models such as those from OpenAI or Anthropic [2]. Their competitive advantage is razor-thin. The underlying technology is not theirs, and the functionality can often be replicated within weeks.

The defensible position

As the wrapper economy reaches its limits, a different type of AI company is rising to the surface. These are companies that do not start from the technology, but from a deeply rooted problem in a specific sector. Their strength is not a clever prompt, but years of domain expertise. They use AI not as an end in itself, but as a tool to optimise complex, existing processes. And that is 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 optimisation and regulatory expertise. Their algorithm 'Alice' predicts and manages energy flows from solar panels, batteries and charging stations in near real time. The result is that companies extract more capacity from their existing grid connection, which alleviates grid congestion.

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

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

At 10x Teamthe defensible position lies in an 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 saves recruiters four to six hours per day, according to the company [4]. But more interesting is what they are trying to solve: bias. Traditional interviews are notoriously subjective. One recruiter pays attention to different things than another. 10x claims objective scoring by applying consistent criteria to all candidates. Their system scores on execution ability, resilience under stress and problem-solving thinking, not on likability or 'cultural fit' as it is often assessed in a first interview.

Angelique Schouten, CEO of 10x Team, emphasises that perspective. "It's not about replacing people. It's about enhancing human potential. That is the future we are moving towards. With AI, not before or by AI." [5]

The fact that 67% of candidates complete 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 job interview looks like and how to scale it without losing quality.

The common thread? These are all companies deploying AI where human expertise is scarce, expensive or inconsistent. They are not automating the simple, but the complex. And that complexity requires domain knowledge you cannot pull from an API. That is their protection against competition.

The next phase

The question is what happens next. Remy Gieling, founder of AI.nl, predicts consolidation. "Over time, many small companies will be absorbed into larger ones," he says [1]. That is logical. The market for 'straightforward' AI assignments is becoming saturated. Companies are looking for specialised knowledge that aligns with their specific needs. That requires capital and highly educated technical staff. The sole traders (67% of current AI companies) are often unable to make that leap.

But another movement is also underway. Companies with genuine domain expertise 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 is the next step: from solving a specific problem to building an ecosystem in which multiple problems within the same sector are addressed. The barrier for competitors becomes broader, deeper and harder to circumvent.

The Dutch advantage

The strength of the Dutch ecosystem lies in the pragmatic application of technology in places where it genuinely adds value. It is a culture where specialist knowledge and an engineering mindset carry more weight than hype. The AI wave acts as a filter. The companies that remain are not those with the most polished interface, but those with the deepest roots in a domain. The future of Dutch AI will not be determined by who can best 'wrap' a language model, but by who can solve the most complex, sector-specific problems. The real protection against competition is not technology you buy in, but the expertise you build up over years. And that expertise, combined with the right technology, is what endures when the hype is over.