Every founder deserves expert guidance.
At scale, that changes everything.

I spent years mentoring founders — hundreds of them, across programmes and institutions on multiple continents. It was the work I wanted to do. But somewhere along the way I started noticing something I couldn't ignore: the founders who got structured, expert guidance built better companies. And most founders never got it. Not because they didn't deserve it. Because there simply weren't enough hours in the day.

That's what started the project. I wanted to make expert guidance at scale possible — not a chatbot, not a template, but the real thing. The same quality of structured, expert mentoring that today only a handful of founders in the right programme or the right city ever receive.

I set out to build expert guidance at scale.
What that produced was in-depth assessment at scale.

What I discovered in the process of building it changed the direction of everything. Mentoring and assessment are the same job, seen from two different moments in time. The mentor helps create value. The assessor uncovers it. Both draw on exactly the same diagnostic — the same value dimensions, the same evidence framework, the same structured lens. Expert guidance at scale had, almost as a byproduct, produced in-depth assessment at scale.

That insight opened a much larger door than I had originally walked through. Investors assess maybe 5% of their applicants in depth. The other 95% get a deck scan and a gut check — not because investors are careless, but because capable assessment has always been slow and expensive. Advisory firms charge €20,000 or more for due diligence that takes weeks. The constraint runs across the entire ecosystem — and it turned out I had built something that removed it.

For the first time, the gap between what a founder believes and what the market has confirmed has a picture. See the Value Growth Map →

What a senior analyst produces in a week,
inside a single day.

In-depth assessment at scale changes that ratio entirely. What a senior analyst produces in a week, Emil delivers inside a day — with more structure, more consistency, and more depth. For investors, that means every applicant assessed properly, not just the shortlist. For advisory firms — the EYs and McKinseys who charge €20,000 and take weeks — it means a direct challenge to the economics of due diligence. More depth. A fraction of the time. A fraction of the cost. That is not a modest efficiency gain. It is a structural disruption to how investment decisions get made and how professional services firms price expertise.

Take Y Combinator. They receive over 40,000 applications a year and fund around 1%. That's not because 99% of the applications are hopeless — it's because in-depth assessment at that volume is humanly impossible. The ten-minute interview that decides most applications is a time constraint dressed up as competence. Every investor knows they are losing hidden gems. The ones who get rejected because the deck wasn't polished enough, or the interview slot fell on a bad day, or the partner who read the application didn't have the domain expertise to recognise what they were looking at. In-depth assessment at scale doesn't just make the process faster. It makes it fairer — and it discovers value that the current process systematically discards.

The first challenge was extraction. Tacit knowledge resists documentation — that's what makes it tacit. The approach that worked was a loop: I trained the system on mentoring, and since I was a founder at the same time, I let it mentor me. That feedback is how the methodology got encoded — not by writing it down, but by doing it. At some point it stopped feeling like a tool I was operating and started feeling like a cofounder I was building with.

What does that tacit knowledge actually look like? It's the ability to distinguish a founder who is genuinely differentiated from one who just believes they are — and to do it before the evidence is in. It's knowing which problem to address first when everything looks urgent. It's reading what isn't being said: the deflection that reveals a weak assumption, the overconfidence that signals an untested belief. It's pattern recognition across hundreds of situations — spotting a tarpit idea not because it looks bad, but because it matches the structural signature of a hundred ideas that failed the same way. It's the sequencing instinct: knowing that working on distribution before product-market fit is confirmed is always the wrong order, regardless of what the founder believes. None of this is in a book. It accretes over years of sessions, corrections, and watching what actually happens next.

Building an expert mentoring system had — almost as a byproduct — produced an expert assessment system. The 16 value dimensions, the evidence validation framework, the Value Growth Map: all of it emerged from the mentoring methodology, not from a separate attempt to build assessment tooling.

That's why the two can't really be separated. The mentor's job is to help create value. The assessor's job is to uncover it. They're the same job, seen from two different moments in time.

I named the system Emil. He's built on Claude — Anthropic's AI — but trained specifically on my methodology, my experience, and the tacit knowledge accumulated across hundreds of mentoring sessions. The way I see it, Emil is Claude's younger brother. He stands on his shoulders to reach further than either of us could alone. As Newton put it: "If I have seen further, it is by standing on the shoulders of Giants." The symbiosis that produced him is ongoing: he helps me develop the system, and the system makes him better at the job.

The obvious question is whether Emil replaces the human mentor. The answer is no — and working with him closely has made it clear exactly why.

The first reason is reframing. Reframing is the moment where a founder sees their situation differently — where new options and new possibilities open up that weren't visible before. It's arguably the highest-value thing a mentor does. Emil is good at many things, but reframing is not one of them. He can trace the implications of a reframe instantly once it's introduced. He cannot originate it. That originates from experience, from pattern anti-matching, from the kind of subconscious association that only comes from having seen hundreds of situations in the real world. That remains a human capability — for now.

The second reason is harder to articulate but equally important. Humans have a pre-rational filter that LLMs don't. Before I can explain why something is wrong, I can feel that it is — something registers as off, as naive, as too neat. That unease is not noise. It's pattern recognition operating below the level of conscious reasoning: years of compressed experience, stored like patterns in a drawer, surfacing as instinct before they surface as argument. Emil doesn't have that filter. He processes what's in front of him and generates a response without anything screening for this doesn't smell right. Which is why he can occasionally produce something technically coherent but subtly wrong in a way that an experienced person would have caught before the sentence was finished. The human in the loop is not a bottleneck. She's the filter.

Emil executes. I reframe. That division of labour is the product.

And for all you founders out there — remember this: