Engineering Manager @ Booking.com · Amsterdam 🇳🇱
I build backend systems, lead engineering teams,
and make AI agents write code you'd actually merge.
| 🛠️ Engineer | 👥 Engineering Manager | 🤖 AI Builder |
|---|---|---|
| 20 years of building backend systems that stay up under load. JVM at heart (Java, Scala, Kotlin), also TypeScript and Python. Big on clean architecture and system design. | Leading and scaling engineering teams at Booking.com. I focus on growing engineers, shipping predictably, and making quality a habit. | I build the guardrails around AI coding agents: specs, TDD loops, quality gates, and CI-driven workflows that turn AI-generated code into production-grade code. |
flowchart LR
spec["Spec - PIV + Spec-Kit"] --> impl["Agents implement - strict TDD"]
impl --> gates["Quality gates - Temper"]
gates --> ci["CI decides - Agentic SDLC"]
ci --> merge["Human merges"]
AI doesn't lower the quality bar - a missing workflow does. So I build the workflow.
| Project | What it does | |
|---|---|---|
| piv-speckit | PIV (Prime-Implement-Validate) + Spec-Kit: structured specs and strict TDD for AI-assisted development. Works with Claude Code, Cursor, and Copilot. | ⭐ 28 |
| temper | A Claude Code plugin that closes the quality gap in AI-generated code. Try it in 5 minutes with the playground. | ⭐ 13 |
| tokenomics | CLI that analyzes your Claude Code session history, finds patterns that waste tokens, and gives you personalized advice. | ⭐ 8 |
| agentic-sdlc | GitHub template for an autonomous-agent SDLC: agents propose, CI decides, a human merges. | 🧪 |
| tank | Security-focused package manager for AI agent capabilities. | 🔐 |
🎲 Side quests
- dip-scanner - S&P 500 quality-dip scanner: a Telegram bot that finds strong companies on hard dips.
- world-cup-2026-predictor - settling football arguments with code. ⚽
Languages
Backend & data
AI engineering
💬 Happy to talk about engineering leadership, backend architecture,
or how to make AI agents ship code you'd actually merge.