MSc Artificial Intelligence student, Utrecht University
Hi, I'm Ruben — I build AI systems that hold up outside the benchmark.
I'm a Master's student in Artificial Intelligence at Utrecht University, researching federated learning over heterogeneous streaming data. Alongside that I build LLM agents, RAG systems, and practical software — and spend a lot of time teaching other people how to do the same.
status = "MSc AI, Utrecht University"
thesis = "federated learning, streaming data"
focus = ["LLM agents", "RAG", "reliable AI"]
teaching = 200 # hours, 8+ courses
Featured work
A few things I've built
From an ontology-aware LLM agent to my current federated learning thesis research.
What I care about
Research interests & focus areas
Federated & streaming learning
Training under non-IID, partially labelled, constantly-arriving data — the focus of my thesis.
AI agents & multi-agent systems
LLM-driven agents that reason over structured knowledge and act reliably, not just fluently.
Neuro-symbolic AI
Combining symbolic reasoning (ontologies, logic) with neural generation for consistency.
Reliable & responsible AI
Evaluating language models honestly, and building systems that fail predictably.
Edge & distributed ML
Machine learning that has to operate under real constraints, not ideal benchmark conditions.
Teaching & mentoring
200+ hours as a TA across 8+ CS courses — I like explaining hard things clearly.
Working on something interesting?
I'm looking for a part-time role (14–20 hrs/week) in AI or software engineering alongside my Master's — or just enjoy talking about federated learning, agents, and reliable AI systems.