Most of the value is in the boring half
Data access, ownership, review workflow, what happens when it is wrong. The model is rarely the hard part and almost never the expensive one.
About
I'm Luca Widmer, an independent AI consultant and coach based in Zürich. I work with teams who are past the demo stage and now have to decide what to actually build — and what to drop.
Before going independent I led a data platform team at a Swiss insurer, where I watched three machine-learning projects succeed and rather more than three quietly get shelved. The successful ones were not the ones with the best models. They were the ones where somebody had done the unglamorous work of deciding who owned the output and what happened when it was wrong.
That is most of what I do now. I come in for a few weeks, help a team find the one or two things worth building, build one of them with their engineers, and then coach whoever has been made responsible for it until they no longer need the calls.
I started out in a Swiss apprenticeship, which means I learned the trade from people who took teaching seriously. That habit stuck. A good share of my work now is the standing hour with someone who has just had "AI" added to their job title and needs to think out loud with somebody who has made the mistakes already.
I work in German and English, mostly with teams between 20 and 500 people, mostly in Switzerland and the wider EU. I do not resell anything and I take no vendor commission.
How I work
Data access, ownership, review workflow, what happens when it is wrong. The model is rarely the hard part and almost never the expensive one.
I would rather ship something narrower that fits how people already work than something broad that everybody routes around.
If you cannot describe what a good output looks like, you are not ready to build. Getting a team to agree on that is often the whole engagement.
The goal is a team that no longer needs me. If an engagement keeps renewing, something has gone wrong with it.
Ask me on a call. It's a better use of both our time than a longer page about me.