A short bio, an honest one.

How I got here

I grew up in Switzerland and studied at the University of Fribourg, where I earned a Master's in Business Informatics. The program sat at the intersection of technology and strategy, which is honestly where I've always been most comfortable — not purely in the weeds, not purely in the boardroom, but translating between the two.

My first real engineering job was at SBB, the Swiss Federal Railways, building safety-critical software that railway workers relied on in the field. That experience shaped how I think about building things: if your code fails, someone's work stops. That standard has stuck with me.

In 2017 I co-founded DataBerg, a web studio with offices across three countries. A year later, I co-founded databaum, an agritech startup that combined IoT sensors and ML models to help farmers manage crop diseases. We onboarded 100 customers in our first season — and I learned more from the things that didn't go according to plan than from the things that did. Building a company from scratch teaches you which problems are worth solving and which are just interesting.

I'd rather ship something imperfect that solves a real problem than polish a solution looking for one.

What I'm building now

Today I build AI-powered systems at production scale — currently processing over 200 million LLM tokens a month. I'm not interested in AI demos; I care about systems that run reliably, that people actually use, and that make a measurable difference.

Recently I shipped an MCP (Model Context Protocol) server — Anthropic's open protocol for connecting LLMs to external tools and data sources. It's the kind of infrastructure work that's invisible when it's done well and painful when it's not. I also work extensively with agent architectures, and I spend a lot of time thinking about the gap between what AI can do and what it should do.

Outside the screen

Since 2017 I've been a part-time instructor at ICT-Berufsbildung Zentralschweiz, teaching everything from mobile development to machine learning. Teaching is the most rewarding thing I do — there's nothing like watching someone grasp a concept that seemed impossible an hour ago.

When I'm not at a keyboard, I'm usually on the water. I earned my RYA Day Skipper certificate in 2025, and sailing is the clearest contrast to my daily work: no abstractions, no deployments — just wind, water, and paying attention.

I'm trilingual (German, English, French), and I've been steadily deepening my knowledge of finance through coursework at the University of St. Gallen and the University of Zurich. I find the same satisfaction in understanding a balance sheet that I do in debugging a distributed system — it's all just finding the signal in the noise.