2021 – now
Grid analytics at TenneT
ML systems and data pipelines for the Dutch–German high-voltage grid: forecasting, anomaly detection and the plumbing that keeps them honest.
Turning turbulent data into things that work.
About
I'm a data scientist and engineer at TenneT, the Dutch–German grid operator, with a PhD in experimental physics from the University of Twente — turbulence, bubbles, and a lot of laser light. That is where the habit of measuring things properly comes from.
Fifteen-plus years of Python later, I build end-to-end machine-learning systems: the model, the pipeline around it, and the boring parts that make it survive contact with production. Lately that means LLMs, and Rust when Python is too slow.
I write about it on the blog, long articles and short TIL notes alike, and I spend as much time as I can underwater.
Things I like
Cold Dutch water, French caves, a scooter when the line is long enough. Every dive is logged — the boring ones too.
Last dive: 8 Aug 2026 · Slijk-Ewijk
Work
2021 – now
ML systems and data pipelines for the Dutch–German high-voltage grid: forecasting, anomaly detection and the plumbing that keeps them honest.
2020
A Jupyter widget for fast, keyboard-driven labelling of text and images. Used by data teams that would rather label than build labelling tools.
2020
A 40-part series on Towards Data Science taking readers from zero to comfortable Python, ten minutes at a time.
2011 – 2016
Experimental fluid dynamics at the University of Twente: bubbles, boundary layers and a lot of laser light. Where the data habit started.
Surface interval
Interesting problem, a talk, a dive trip? open to interesting projects — the fastest way is e-mail.
Last dive: 8 Aug 2026 · Slijk-Ewijk