I build tools that make complex things clearer, calmer, and more useful.

AI Engineer · Product Builder · Berlin / remote
Portfolio · CV · LinkedIn · Culture

I keep noticing systems people have learned to tolerate: rights that are difficult to understand, public services hidden behind forms, important information scattered across institutions, and software that creates more work instead of less.

My instinct is usually the same:

Why does this have to be so complicated? What would the better version look like?

So I build prototypes of that better version, put them in front of people, and learn where reality disagrees.

Technology should return human attention to things worthy of being human.

Make legal rights easier to understand and act on.

A live research system across 5,936 German federal laws. The Mietrecht pilot takes plain-language questions, finds relevant BGB sources, keeps uncertainty visible and turns user-found failures into regression tests.

→ Try GitLaw

Make public changes easier for citizens to notice and understand.

Open-source watchdog agents monitor public sources, write cited digests and machine-readable logs, and keep a human review step before publication.

→ Open the public portal

Automate repeatable work without quietly automating responsibility.

A reusable architecture for AI workers with explicit evidence, permissions, human approval, observable traces and replayable failures. Current validation is synthetic; the next proof is a real workflow in shadow mode.

→ Open technical demo

Let skilled reviewers spend attention on judgement, not reconstructing documents.

A public-sector Document AI prototype that connects documents, versioned rules and evidence gaps into human-reviewable next steps. Public cases use synthetic data.

→ Try PrüfPilot

Human problem → useful workflow → grounded AI → visible uncertainty → human control → real feedback → regression tests.

I care especially about systems where mistakes have consequences: law, public services, care, health and operational work.

Python · FastAPI · TypeScript · React · LLMs · RAG · AI agents · MCP · structured outputs · evaluations · human-in-the-loop

The four projects above are the clearest proof of my current direction. The wider archive includes SafeVoice, civic infrastructure, human-systems experiments, peace education, product ideas and games.

→ Explore the project map
→ See the human layer

A strong team where I can own difficult problems end to end — understand the human need, build the system, test it in reality and improve it.

Applied AI · AI Engineering · Product Engineering · Berlin / EU / remote

Let’s talk