Short note on who I am, what I build, and the angle I take on AI products.
I'm an Applied AI & Product Engineer — 20 years building products, now building the systems behind them.
For two decades I shipped software as a product manager — HelloFresh, Native Instruments, Pivotal Labs — operating at scale for 8M+ users. Then I did the thing I'd always meant to: 1000+ hours of software and AI engineering training, and crossed over into building.
Today I design and ship AI systems end-to-end: tool-calling agents, RAG, evals, and the observability that tells you whether a model output became a feature or a bug report. At Liv Longevity Labs I replaced a fragile no-code workflow with a production LangChain assistant in three weeks.
What's hard to hire for: senior product judgment on every engineering decision. I can talk trade-offs with an exec in the morning and ship a reliable pipeline in the afternoon.
Based in Berlin, available remotely. Open to roles where hands-on engineering meets product thinking.
Click a role to expand the detail.
Designed and shipped a LangChain tool-calling agent and health assistant, replacing a fragile no-code (n8n) workflow with a production-grade, code-based system — in three weeks.
Built the AI observability layer with LangSmith for data-driven iteration, drove architecture decisions, and improved instrumentation. Personalized longevity tools on top of epigenetic biomarker data.
Stack: LangChain · LangSmith · Next.js 15 · Supabase · TypeScript.
Agent design, integration with existing tools, and AI workflow design for small businesses that need to move fast — hands-on engineering, not slideware.
Led the modernization of HelloFresh Group's Delivery Management system — the data infrastructure and business logic powering customer-facing experiences for 7M+ active customers across 7 brands in 22 countries.
Owned the roadmap and prioritization, balancing high-impact delivery against stability across a long modernization arc. Drove a company-wide AI translation/localization strategy: discovery, tooling, integrations, and validation of a scalable platform across brands and countries.
Owned the roadmap, development, and launch of Maschine+, NI's first standalone music-making groovebox — exceeding sales targets within three months and selling out main shops. Cracked IoT usability problems that had been launch blockers, getting a previously-cancelled project greenlit.
Led the modernized release of Native Access, NI's most-used product (650k+ users), driving UX and the technical roadmap across partner software teams.
Hands-on PM consulting across startups to enterprises — Volkswagen, Audi, Bosch, the Dutch Ministry of Infrastructure, ReDI School. Directed end-to-end development and coached teams on roadmapping, story mapping, prioritization, and continuous delivery.
Shipped across IoT and connected devices, APIs, auth systems, mobile apps, and internal tooling — pairing directly with engineering and UX. Back then the transformation was Agile; the shape of the work is the same as AI today.
Joined as one of the first employees and helped build the processes and product thinking that scaled eduK to 9 million users and 3,200+ courses — landing it on Fast Company's Most Creative Companies in Latin America.
Built scalable market-research methods for category expansion, ran international content during LatAm expansion, and sat on the executive team.
A snapshot. The bars are honest-ish and re-calibrated quarterly.
The switch into engineering, on the record — two bootcamps, an MIT course, and earlier degrees.
Things I do that aren't code. They all show up in the code, eventually.
Building something where AI meets product? Let's talk.
Whether it's a startup's first version or a new bet inside a larger org — turn a thesis into a product real users actually touch.
You have a product and a hunch that AI could help. I'll find where LLMs create real value, build the case, and ship it reliably.
Senior product judgment on every engineering decision. I talk trade-offs with an exec in the morning and ship a pipeline in the afternoon.
A small team that needs AI features fast and hands-on, without a full-time hire.