I help my customers embrace AI, with solutions built around how they actually work.
Every AI decision is a set of trade-offs — what it costs to run, how well it performs, how much work it genuinely saves, and whether it holds up when someone checks. I have spent nine years working where those trade-offs get settled.
Where I think this is heading: Physical AI — models that act in the world rather than only produce text. As the models themselves become commodities, the difficulty moves to everything around them: the compute budget on the device, the latency, and the safety case for a machine that moves. That is the ground I have spent my career on.
Nine years in automotive semiconductors, making models run economically on real silicon — and a granted patent in AI hardware safety. I work through the cost implications of different architectures with clients, before anything gets built.
Agents, retrieval and LLM systems built, deployed and handed over. We go through the work together to find which parts are genuinely worth automating — and which are better left to people.
At AUMOVIO I co-authored the company's first internal AI governance process, aligned to ISO/PAS 8800 and the EU AI Act. I bring that in early and work alongside your legal and technical people, so approval keeps pace with the build.
What I offer.
Build custom solutions, or get a technical deep-dive.
Custom AI solutions
Agents and systems designed with the people who will use them. Built, deployed and handed over as something your team owns.
- Agents for repeating workflows
- Retrieval over your own material
- Document and data extraction
- Deployment and handover
Compliance and safety
Worked through with your legal and technical people from the start, so approval keeps pace with the build. Also available as advice on its own — reviewing what you already have, or sitting alongside your team while they build it themselves.
- EU AI Act classification
- ISO/PAS 8800 · ISO 26262
- GDPR basis and transparency
- Internal governance processes
- Consultancy support and guidance
AI business case evaluation
What a solution would cost to run, what it would do, and whether those justify each other — before anything is built.
- Cost levers, honestly modelled
- Expected functionality
- Build, buy or leave it
- A case you can take to a board
LexTalent AI — in testing with its first client.
A sourcing agent for an executive search firm placing lawyers. Work that used to take weeks of manual searching is now a single button press. The guardrails are the part I care most about: catching artefacts from AI hallucinations, while making sure no valid result is missed.
More about itReach out if you're curious whether AI fits your daily work.
A sentence or two about what you do is plenty to start.
Frankfurt am Main
English · Deutsch