Product strategy & vision
Two-year visions, Jobs-to-Be-Done, competitive benchmarking, future press releases. Narratives that align engineering, sales and leadership on one direction.
8+ years across B2B and B2C SaaS, turning operationally messy problems into shipped product and measurable outcomes. Most recently as an AI product manager, taking a multi-agent LLM platform into production in a high-liability regulated domain.
Strategy and execution end-to-end — vision, discovery, roadmap, pricing — across relocation, rental operations and automotive electronics.
Six things, one of which is AI. The other five are what make the AI work land.
Two-year visions, Jobs-to-Be-Done, competitive benchmarking, future press releases. Narratives that align engineering, sales and leadership on one direction.
LLM agent orchestration, RAG and retrieval design, prompt versioning, evaluation with golden datasets and LLM-as-judge, human-in-the-loop rollout. Production, not demos.
Rate engines, licensing models and tier design. The Rentman licensing and pricing rework returned 40% revenue growth on its own.
Turning fragmented operational knowledge — spreadsheets, inboxes, tribal expertise — into product infrastructure that scales without more headcount.
100+ customer interviews, design sprints, dogfooding my own products. Deep research that changes roadmaps rather than decorating them.
Analytics frameworks, success metrics and transparent roadmaps in Mixpanel and SQL. Every decision tied back to a business number.
A multi-agent system that turns relocation expert knowledge into platform capability — retrieval, evaluation and human-in-the-loop.
Centralised vendor communication that cut response times in half and ended inbox fragmentation.
Repositioned a mid-market scheduling tool as an enterprise operations backbone — 80% revenue growth.
A flexible rate engine that unlocked U.S. market entry — 25% more license upgrades.
Normalized edit distance runs on every interaction — the passive quality signal that needs no reviewer.
“Normalized edit distance is a passive quality signal on every single interaction — no reviewer required.”
What I actually work in, grouped by how deep it goes — and the courses that put the AI half on a footing.
Shipped to production, with the evals to prove it
Enough to argue with engineering, never enough to do their job
Daily drivers across eight years and four product teams
Certified Scrum Product Owner (PSPO I)
Scrum.org
Certified Scrum Master (PSM)
Scrum.org
Certified AI Product Manager
Product HQ
AI Evals for Engineers & PMs
Maven · 2025 ↗
Agentic AI for Product Managers
Maven · 2026 ↗
Retrieval Augmented Generation
DeepLearning.AI · 2025 ↗
Eight years across two of the Netherlands' fastest-growing SaaS companies took me from product specialist to senior PM — B2B ERP at Rentman, then relocation software at Settly. Along the way I built the habit that still shapes the work: dogfood the product, talk to a hundred customers, and only then draw the roadmap.
Lately I've gone deep on AI product — multi-agent architectures, retrieval design, and the evaluation frameworks that decide whether any of it can be trusted in a regulated domain. It's the most interesting problem I've worked on. It is not the only thing I do.
I made most of the mistakes already — bad roadmaps, shipping the wrong thing, misreading enterprise needs. That's what makes the work interesting now.
“The metric first, then the mechanism, then the trade-off. Discovery before roadmap; evaluation before scale. Comfortable in the code enough to argue with engineering, never enough to do their job.”