Cloud, platform, and automation engineer. I build AWS infrastructure, CI/CD, and agentic-AI tooling for software that has to pass an audit — most recently at Philips in Eindhoven.
I work where cloud infrastructure meets software that has to be provably correct. At Philips that meant the platform's first AWS WAF delivered as infrastructure-as-code, CI/CD pipelines that hold up through a release, and design documentation that survives a PDLM/QMS audit.
Alongside that I build with Claude, MCP servers, GitHub Copilot and n8n — and I've driven adoption of those workflows among senior colleagues, not just used them myself.
An MCP server that exposes the Dutch IND recognised-sponsor register to AI agents, so an assistant can answer "is this employer a recognised sponsor?" against the live register instead of guessing. TypeScript, published as a Model Context Protocol server.
View source on GitHub →Extended BATDOK so electronic medical records stay available where the network doesn't: IPFS-backed peer-to-peer sharing via an Android plugin bridging Kotlin and Go over JNI, with a feasibility and security assessment for military-grade deployment. Built during my Philips internship.
Related public code: gomobile-cal →The platform's first WAF, delivered as infrastructure-as-code: OWASP Top 10 rules, IP filtering, DDoS mitigation and automated rule updates, with CloudWatch logging and alerting for the security signal. Delivered at Philips.
Enterprise work — no public code.
Deployment automation for a self-hosted n8n instance on Fly.io — reproducible automation infrastructure you own, rather than a SaaS tier.
View source on GitHub →This site: Astro 7 on Cloudflare Workers, with a Sveltia CMS and owner-only GitHub OAuth —source on GitHub.
Happy to talk about cloud security, CI/CD in regulated environments, or where agentic tooling actually holds up in production.