I build AI systems that work past the demo.
Software engineer based in Geneva. I focus on AI automation, agent infrastructure, and practical tooling — the boring, load-bearing layer that makes smart models actually useful in production.
Quick facts
Background
I've built production systems across data infrastructure, health tech, and AI tooling. My focus has shifted steadily toward agent-based workflows — not because agents are fashionable, but because the hard problem of making AI reliable in complex environments is genuinely interesting.
I care about the operational layer: eval pipelines, resumable runners, multi-model orchestration, and the debugging you have to do when things break in ways that aren't in any tutorial.
How I work
I'm a systems thinker. I reach for first principles when something doesn't add up, and I'd rather build the mechanism than write a rule about it.
I work independently — small teams, direct collaboration, no extra process. If you're building something at the intersection of AI and real workflows, that's where I'm most useful.
Experience
Built data pipelines and tooling for clinical genomics workflows at scale. Led engineering work at the intersection of precision medicine and infrastructure.
Designing and shipping personal automation systems: multi-model orchestration, voice tooling, research acceleration pipelines, and agent infrastructure for real workflows.