Sharing lived experiences from building and implementing AI.

I started building things before I was qualified to. At sixteen, I was running a small business out of my bedroom shipping night-vision equipment — handling the product copy, customer service, logistics, all of it, because there was no one else. Later, while I was studying medicine, I created a card game for medical students and doctors that went on to sell thousands of copies, win an award, and get covered in the press. I didn't think of any of this as entrepreneurship at the time. I was just following what I was curious about. The pattern only becomes obvious in retrospect: I have always been building things. I just started playing for higher and higher stakes over the past decade.

At twenty I served as an airborne infantry reserve officer cadet in the German Bundeswehr's Division Spezielle Operationen and graduated best in class — leading forty soldiers under field conditions for the first time. Medical training in Saarland followed, with a Master of Science in Public Health at the London School of Hygiene and Tropical Medicine completed in parallel. Later came a Master of Public Policy at Oxford and a Postgraduate Diploma in Entrepreneurship at the University of Cambridge.

None of these were strategically planned. I had been taught, like most people with a strong undergraduate record, to stay focused on one topic and go deep. I did the opposite. I had broad curiosities and a stubborn preference for doing things at a world-class level rather than at a domain-appropriate one. At the time, this looked unfocused to most of the people around me. It is only now — when the work that compounds is increasingly the work of connecting domains that don't obviously belong together — that the breadth has turned out to be a real asset. I didn't plan that. I just kept following what I was interested in.

After the degrees, I joined the founding team of smart Helios, the first digital health venture inside the Helios/Fresenius hospital group. From there I went to FoundersLane as Equity Partner and Chief Medical Officer, helped build it into one of Europe's larger corporate venture builders, and stayed through its acquisition by Creative Dock in 2022. Along the way I advised companies across the European digital-health ecosystem, including Wellster Healthtech, where the work contributed to a nine-figure outcome.

Then came Halitus. I co-founded a medical diagnostics company, raised a million euros in pre-seed including from Roche via RoX Health, and spent three and a half years on it. The market did not develop fast enough for the technology, and in 2025 I wound the company down — after selling an apartment to extend its runway past what the market was prepared to give. It was painful. It cost me money, time, and a stretch of real self-doubt. It also taught me more about building, financing, and timing a regulated business than any of the successful chapters did. That period has turned out to be the most useful education I have had as an operator. The lessons are now visible in everything I do.

Aiomics is the proof. We build clinical intelligence for European hospital networks — software that keeps patient data correct as hospitals let AI take on more of the documentation load. It runs under GDPR and ordinary hospital data governance. At pre-seed stage we are already piloting across some of the largest hospital groups in three countries, with master service agreements on track to reach millions in ARR. The traction is partly because the problem we are solving is large and clearly felt — and partly because I am building this company with the lessons of Halitus in the room. We have a strong international team. We are signing customers most companies do not see until Series A. The contrast with the previous chapter is not lost on me.

Alongside Aiomics, I have started angel investing, with early results that have surprised me on the upside. I advise leading investors including Speedinvest, and I work with private equity firms and M&A advisories on healthcare and AI deals.

I write, I speak, and I keep teaching, because explaining what I am learning is how I make sure I am actually learning it.

Two things I wish I had clarity on sooner.

Two patterns have held across all of it.

The first is one I came to late. For most of my career I focused on exactly the kind of work that AI is now starting to do well — execution, analysis, the visible technical layer. It was what I had been trained to value, and what I was rewarded for valuing. Only in the entrepreneurial chapter did I start to understand how much of the actual leverage lives in the parts I had been ignoring: framing the problem, choosing the team, building the conditions for good judgment, deciding what reaches the customer. The recognition came slowly and partly under duress, which is probably the only way it ever comes. It is also the recognition that AI is now forcing on everyone, on a faster clock.

The second is that staying close to the details, longer than seems reasonable, is the single most consistent compounding move I have seen — across surgery, the military, founding teams, advising, and writing. The founders who keep walking the factory floor after they no longer need to. The clinicians who keep seeing patients after the title says they shouldn't have to. The investors who still read the primary documents. It is a slower path on most days and a much faster one over years. I work that way deliberately, and most of what I argue for in public is some version of that argument.