Wouter Durnez
AI & Software Engineer · Ghent, Belgium
wouter@rugvin.be · GitHub · LinkedIn · HF · ORCID · Scholar · DBLP
Summary
AI & software engineer building LLM tooling and agent infrastructure. PhD in experimental psychology, MSc in AI. Based in Ghent, Belgium.
Experience
-
Founding AI Engineer
Voice-AI agents for blue-collar recruitment — agent infrastructure, MLOps around phone-call workloads, 22-language coverage.
-
Senior → Staff AI Engineer
GenAI platform: RAG, asset summaries, AI Page Generation, tool-wielding conversational agent in the flagship product.
-
Senior Data Scientist
Internal HR recommender — fuzzy skill extraction, weighted collaborative filtering, embeddings-based conversational variant.
-
Python Developer
Flask-based BI template — auth, structured logging, CI/CD, Docker.
-
Senior Researcher
HCI research with biometrics and smartphone logging, plus the ETL/analysis frameworks that made it tractable.
Education
-
MSc Artificial Intelligence (Engineering & Computer Science)
-
PhD Experimental Psychology
-
MSc Theoretical & Experimental Psychology
Skills
- Languages
- Python, TypeScript, SQL
- AI
- LLM tooling, agent infrastructure, RAG, prompt engineering, tool use, evals, voice-AI agents, recommender systems, deep learning
- Infra
- AWS, Docker, CI/CD, structured logging, MLOps
- Libraries
- Fence (author), Astro, Flask, PyTorch
- Research-adjacent
- HCI, biometrics / eye tracking, experimental design, statistics
Open source
Full catalogue at /projects. Marquee: Fence — lightweight LLM & agent library, MIT, ~2.5k monthly PyPI downloads.