About

Senior Researcher in Safe and Trustworthy AI in Proteomics at Utrecht University, affiliated with the AI Technology for Life and the Biomolecular Mass Spectrometry and Proteomics groups. I work on safe, trustworthy AI: at present this centers on developing efficient foundation models for mass spectrometry while integrating principled epistemic uncertainty to ensure reliability in downstream scientific and clinical interpretation.

I have a broad, hands-on background spanning symbolic and data-driven AI, and I have applied AI methods across diverse domains including logistics, transport, software engineering, health, and bioinformatics. My research consistently focuses on making learning more efficient (optimization, compact architectures) and making model behavior more transparent. I have developed axiomatic and game-theoretic explanation frameworks—including Shapley-based and interaction-aware formulations—for a range of machine learning models, aiming for explanations that are computationally tractable yet faithful. In decision-making, I study preference modeling under uncertainty, designing methods that aggregate, rank, and analyze criteria and stakeholder inputs in probabilistic and robust ways.

I earned my Ph.D. from TU Delft with cum laude honors (the highest distinction in the Netherlands). I then held postdoctoral positions at JADS / Eindhoven University of Technology and in the Learning and Reasoning group at Vrije Universiteit Amsterdam. I have also conducted research visits to the Rational Intelligence Lab at CISPA (March–June 2025) and the Department of Mathematics and Informatics at the University of Marburg.