For my thesis at UGent, I trained four real-time speech enhancement models in PyTorch. The best was a transformer-GRU hybrid, which I shipped as a streaming ONNX model that runs live in the browser.
I measure every change on the full test set and build agentic workflows with Claude Code, sub-agents, and MCP servers. My AI projects also include tree measurement using YOLOv8 and multi-view geometry.
Automation was part of my electrical engineering studies, and maintaining ABB robots at Volvo Cars showed me what software looks like when it moves real hardware. I’ve also worked as an Automation Engineer at Van Heyghen Staal and a Digital Manufacturing Engineer at Capgemini.

