If you want a real AI ecosystem, builders need a place to build

The Dutch AI scene has no shortage of events, networks, and ambitious claims. What it often lacks is more space for the people actually building. Philip Gast has made that gap part of his work.

With AdamI, he helps companies move from AI ideas to real products. With The AI Foundry, he is trying to create the kind of technical community he felt was still missing in Amsterdam: less theatre, more substance, and more room for founders and engineers to learn from each other. That same builder mindset also shows up in a joint effort with RankmyAI to launch the Amsterdam AI landscape next month. In conversation, he was clear about one thing: AI becomes interesting when it moves beyond slide decks and starts reshaping real work.

Philip Gast is co-founder of AdamI and founder of The AI Foundry in Amsterdam.

A builder’s mindset, early on

Long before AI became fashionable, Philip had already found the combination that still defines how he works today: technical depth, problem solving, and a strong preference for making things that actually work.

“I actually wanted to become a surgeon,” he says. “But I always had an affinity with computers. I liked puzzles, solving problems, building things.”

That instinct showed up early. He built his first game when he was ten: a simple adventure game, somewhere between Mario and a basic RPG, with a knight moving through a story world he had invented himself. Basic, by his own admission, but that was never really the point. What mattered was the feeling of making something work.

When it came time to choose a study programme, someone suggested artificial intelligence. He barely overthought it. The label appealed to him, but more importantly, the field itself turned out to fit almost perfectly.

“People always say you spend your life searching for your passion. Some people find it at forty, after doing something completely different first. For me, it was almost a direct hit. From the age of nineteen, I knew this was my thing.”

He started in Groningen, drawn by the programme’s technical focus and its more hands-on approach. After five years there, he returned to Amsterdam for his master’s. That route still shows in the way he works now: comfortable with technical depth, but equally at ease connecting ideas, people, and business opportunities.

That mix creates opportunities, but it also creates tension.

“In AI, there is just so much you can do. That’s the fun of it, but also the danger. You have to focus. I’m still learning how to say no.”

From AI ambition to actual products

That tension between exploration and focus comes through clearly in Philip’s work with AdamI. If The AI Foundry is where he builds for the ecosystem, AdamI is where he helps organisations build for themselves.

He is quick to push back on the idea that AdamI is a traditional consultancy.

“We’re not really a consultancy in the classic sense,” he says. “We are a group of slightly crazy developers who want to do cool projects. We always work towards a product, not just a slide deck.”

That distinction matters. AdamI does not just advise on AI strategy. It helps clients scope use cases, shape the business case, and turn those plans into working products, platforms, and proofs of concept. For Philip, that is where AI becomes real.

The company has grown quickly, reaching sixteen people in roughly a year, with five founders in total. That setup gives the team room to move fast, although the familiar startup reality still applies: sales remains a constant challenge.

Looking ahead, he sees AdamI moving beyond one-off project work and towards reusable infrastructure. Many companies are still trying to build their own AI platforms from scratch. In Philip’s view, that means the same mistakes are repeated again and again.

“We now see very clearly what you need to buy, what you need to build, and what you need to watch out for. So we’re moving towards a more complete solution. At some point, that becomes the model: building ideas on top of our platform.”

We always work towards a product, not just a slide deck.

That mindset also shapes the kind of projects he finds most interesting. Not because they use the newest models, but because they change how work gets done.

One example is a project in Suriname, where he helped deliver an AI platform, train people, and identify new use cases on the ground. For Philip, the appeal was not just the technical complexity, but the chance to build something meaningful in a very different context.

“At this age, being able to go there and say, alright, let’s figure this out, that’s just great.”

He is equally energised by projects where AI affects the structure of work itself. 

“I’m usually operating between technical feasibility and business reality.”

That bridging role is important to him. He is less interested in AI as a demo layer than in AI as a practical shift in how organisations operate.

Beyond automation

Philip is careful not to reduce AI to chat interfaces or generic productivity gains. Automation matters, but he sees the bigger opportunity elsewhere.

In his view, organisations tend to approach AI through two tracks. One is improving or automating what already exists. The other is more fundamental: using AI to redefine how a process, service, or proposition can work in the first place.

That second track is where things get interesting.

“When you really see that the way of working changes, those are the great projects. That’s much more interesting than just building another agent.”

That is also why he is drawn to work that goes beyond incremental efficiency. The more compelling question is not whether AI saves a bit of time, but whether it changes what becomes possible.

What the ecosystem is still missing

If AdamI is about helping organisations build, The AI Foundry is about helping builders find each other.

Philip started The AI Foundry because too much of the Dutch AI scene felt stiff, over-structured, and quickly pushed into commercial formats. In his view, that creates plenty of activity, but not always the kind of environment where founders and engineers actually learn from one another.

“I wanted to make the ecosystem less stiff,” he says. “In the Netherlands, things are often very thought through and quickly become commercial. Sponsors want things, and before you know it, you end up with very commercial events.”

What he wanted instead was a place built more around substance than optics. Somewhere technical people could work, meet, and exchange ideas without everything first having to be filtered through sponsorship decks or ecosystem jargon.

The result is The AI Foundry in Amsterdam, where a growing group of AI companies work side by side and where events now attract a recurring community of founders and engineers. Some sessions go deep into technical topics. Others help founders with practical matters around building companies. The shared principle is usefulness.

“It has to stay low-threshold and enjoyable. A place where engineers can work and learn from each other.”

What makes this more than a side project is that the two worlds feed each other. The client work keeps Philip close to the real constraints and opportunities organisations are dealing with. The community work keeps him close to the people building new things from the ground up. Together, they create a feedback loop between practice and ecosystem.

Making builders visible

That same instinct also led to the Amsterdam AI landscape map. Philip started the initiative as a way to make the city’s AI activity more visible and to show that, despite the usual narrative, plenty of people in Amsterdam are already building.

“People can be very negative. We’re behind, we’re late, all of that. But there are already plenty of people here who are just building.”

What began as Philip’s initiative is now being taken to a next level together with RankmyAI. The Amsterdam AI landscape is set to launch on RankmyAI next month, not just as a one-off overview, but as a more dynamic and shareable format that can also be used by other organisations interested in mapping their own ecosystems.

For Philip, that matters. Visibility is not just about promotion. It is a way of strengthening the ecosystem by making builders easier to find, connect, and recognise. In that sense, a map is not just a visual layer. In the right form, it becomes part of the infrastructure of an ecosystem.

Collaboration beats ecosystem theatre

Philip first connected with RankmyAI after meeting David at an event. What stood out to him was not just the idea itself, but the possibility of building something complementary rather than competing by default.

“It’s very Dutch to think: I’ll just do it myself. I’m a bit done with that. Working together is more fun, and it lets you think bigger.”

That attitude says a lot about how he views the ecosystem more broadly. He is not especially interested in multiplying initiatives for the sake of it. He is more interested in what happens when different efforts actually reinforce each other.

That also explains his limited patience for larger structures that spend a great deal of time discussing, aligning, and organising, but struggle to land something tangible. For Philip, a healthy ecosystem is not built through endless coordination alone. It also needs places where things are simply being made.

From platform to infrastructure

Toward the end of our conversation, I asked Philip what he would like RankmyAI to become from his perspective. His answer came quickly.

“The Bloomberg or Reuters of the AI landscape in the world”

He does not mean that as a media vanity position. He means it as a trusted infrastructure role: a platform that helps make sense of the field while becoming stronger through the quality of its data, research, and visibility.

That is why he sees more opportunity in interviews, ecosystem studies, and structured knowledge products than in weaker commercial shortcuts.

“If your business model strengthens the quality of your platform, that’s a very good sign. You have the data. You can do interviews. You can do research. You can make that position stronger.”

He also argues for a more visible and more ambitious role. Show the field. Show the companies. Show what is moving. Not just in Amsterdam, but across the Netherlands.

“Open that map up. Show that you are the party that knows the ecosystem.”

That feels like the right place to end. Philip is building in two directions at once: helping organisations turn AI into working products, and helping the broader ecosystem become more useful for the people actually creating those products. In both cases, the same instinct keeps coming back. Less theatre. More substance. Less ecosystem talk. More room for builders.