Tools I wanted to exist, built in the evenings and pushed to GitHub.
Most of my side projects start as a personal itch and end up public. MCP servers that give AI assistants a way into real services, reactive collections and WebSocket hooks for the Payload ecosystem, and the occasional detour into hardware — like reverse-engineering a Bluetooth LED matrix until it did what I wanted. Nothing grand, just things that work.
The infrastructure side of the job — keeping what we ship boring and reliable.
Alongside the engineering work I look after the platform it all runs on: a Docker Swarm cluster, CI/CD pipelines in Azure DevOps, and the monitoring that flags trouble before anyone notices it. Supporting work by design — done well, nobody has to think about it.
My main role. Full-stack engineering and AI systems that make it into real production use.
I build AI-powered applications end to end — interface, back-end, and the model plumbing in between. At Stofloos that means sitting directly with clients and their users rather than behind a layer of account managers: healthcare providers cutting the administrative load on practitioners, manufacturers turning decades of domain knowledge into something a model can actually reason with. Short lines, full ownership from first sketch to production.
Successfully completed a Bachelor's degree in ICT at Dutch University of Applied Sciences, graduating with highest honours (cum laude).
Designed and built an automated deployment solution for Stofloos as part of my graduation internship, enabling streamlined software rollouts with real-time monitoring.
For my graduation internship, I developed an end-to-end deployment automation platform at Stofloos. The solution provides a user-friendly interface to deploy software applications and modules, with clear status indicators and monitoring capabilities. This reduced deployment overhead significantly and improved reliability across environments. Graduated cum laude based in part on this project.
A semester away from the web stack, listening to machines through Sorama's acoustic hardware.
For my minor I swapped web applications for microphones. Using acoustic camera hardware from Sorama, we worked on detecting machine malfunctions early from the sound they make — the premise being that a failing bearing announces itself long before it stops. Signal processing in Python, a TypeScript front-end to make sense of it, and a real appreciation for how different embedded constraints feel from a web server.
Two years of shipping production code on the days I wasn't in lecture halls.
After my first internship Stofloos kept me on part-time, and I stayed for over two years while finishing my degree. Front-end and back-end work on client applications, increasingly on the AI side as that became the centre of what we do. It ended in May 2024 the way you'd hope — by turning into a full-time engineering role.
My first internship at Stofloos, and the first thing I built that other people relied on.
I designed and built a workflow engine from scratch: teams define a repeatable business process once, then trigger and monitor it without touching it again. PostgreSQL as the data backbone, DevOps practices baked in from the first commit. Six months that moved me from "I can write code" to "I can ship something that runs in production".
Connecting a deliberately awkward IoT device to a real IoT platform, over long-range radio.
OpenRemote challenged us to hook something "wacky" up to their IoT platform. We built GPS trackers that talk to each other over LoRa long-range radio in a mesh network, relaying their positions through an edge device up to their cloud. Written in C++, and my first proper lesson in what happens to software when the network is slow, lossy, and entirely your problem.
Nature & Technology profile
Higher General Secondary Education (HAVO)