About
More than a CV
I'm a Dutch Computer Science and Data Science graduate now doing a Master's in Artificial Intelligence at Utrecht University. Here's the version of my story that doesn't fit on a résumé.
Why AI
What pulled me toward AI wasn't the hype, it was the gap between how machine learning is usually taught — clean datasets, IID assumptions, a fixed test set — and how it actually has to work once it's deployed. My thesis on federated learning over heterogeneous, partially labelled streaming data is basically an attempt to sit inside that gap on purpose: data that's messy, distributed, and never really "done" arriving.
Research, done properly
I like research that stays honest about its own limitations. Comparing LLM-generated first-order logic against a symbolic prover (Prover9) on the FOLIO dataset, or building an ontology-aware agent that checks its own text for logical consistency, appeals to me for the same reason: both projects are about not trusting a model's fluency as a proxy for correctness.
Building practical software
Research interests aside, I like shipping things that work end to end — a RAG chatbot that actually retrieves the right document, an IoT watering system that survives a flaky sensor, a roguelite built solo in Godot because game feel is its own kind of engineering problem. Full-stack, hardware, or purely software, the throughline is finishing the thing.
Continuous learning
I spend a lot of time in new tooling — Claude and its agent ecosystem, n8n, MCP — partly because I'm curious, and partly because 200+ hours as a teaching assistant has made me somewhat obsessive about actually understanding a tool well enough to explain it, not just use it.