Own projects where I build the whole chain — from raw source to a working dashboard or AI application, with the full approach and the code in the open so you can see how I work.
Digital twin: an AI agent that answers for me — with sources
A LangGraph agent over my CV, projects and writing that answers visitors' questions in a streaming chat: every claim cited, refusal when there is no source, a hard daily cost cap, and human approval before it acts on my behalf.
Fail closed, no model
95% tools · 90% tasks
Human approval
Open data warehouse: a Kimball star schema on RDW and CBS
End-to-end ELT platform on Dutch open data: Python ingestion, dbt transformations and a documented star schema — with tests and CI on every pull request.
3 documented grains
Lint + build + tests
One command
Real-time public transport streaming pipeline on Kafka
Every vehicle in Dutch public transport, streamed live — from GTFS-realtime feed to a map of the whole country, with the correctness guarantees to go with it: per-partition watermarks, at-least-once with idempotent upserts, a dead-letter queue and replay from the raw archive.
176 messages/s
5 min · 4 dimensions
DLQ + replay
Strafrecht-RAG: no source, no answer
RAG over 1,906 Dutch criminal-law rulings with one rule: no source, no answer. Every claim ends in an ECLI, section and paragraph — and that reference is checked in code, not requested from the model.
100%, by construction
89%
1,906 rulings · 0 parse errors