PhD Researcher · University of Amsterdam
Human-Centred Explainable Conversational AI
The design problem is not only whether an explanation is understandable or trustworthy — but whether commitment is warranted. Investigating when conversational AI should commit to an answer in uncertainty-prone AI–human dialogue — and when it should ask instead — to support appropriate human reliance.
About
I am a PhD Researcher and Research Scientist at the Hogeschool van Amsterdam, pursuing my doctorate at the University of Amsterdam. My research is in Human-Centred Explainable Conversational AI, with a focus on how AI systems should manage commitment under uncertainty to support appropriate human reliance.
The central problem my thesis addresses is premature commitment: a conversational system can answer before its interpretation of user intent is sufficiently resolved, and a human can accept an AI recommendation before independent evaluation has occurred. As part of a collaborative research programme, my doctoral work investigates the conditions under which conversational explanation delays, supports, or justifies commitment on both sides of this interaction.
My published work investigates how perceived explainability and anthropomorphic design jointly shape trust through social interpretation, and how a Dempster-Shafer belief state can serve as an interrogative gate — blocking answer generation when epistemic uncertainty is too high and prompting the system to seek clarification instead. A complementary behavioural study examines whether independent judgement and cognitive forcing before recommendation exposure improve correctness-sensitive reliance.
I am currently extending this work toward an integrated HC-XCAI reference architecture that combines AI-side interrogative gating, evidence-grounded retrieval, source provenance, and human-side commitment gates — with longer-term interest in interactive reasoning under uncertainty in agentic AI settings.
Published Work
Expertise
Explainable AI
Knowledge Graphs & Retrieval
NLP & Conversational AI
Research Methods
Engineering & Tools
Design & UX Research
Side Projects
Small technical prototypes and research-adjacent tools developed to make ideas tangible, test workflows, and explore practical interfaces for human-centred computing.
Local toolchain · Open source
A local-first LaTeX workspace for editing source files, importing project ZIPs, compiling PDFs with bibliography support, and organizing source files locally. Desktop release builds are available for Linux and Windows.
Contact
I'm always open to conversations about XAI, conversational AI, knowledge graphs, or agentic systems. Reach out for collaborations, speaking invitations, or just to talk research.