Work

Proof beats promises. Four capability areas, each earned the hard way — on client platforms and in products we design, ship and operate ourselves.

Data Strategy

Regulated data platforms, cloud-first

Architecting the first cloud-first data platform of its kind in Belgian federal social security — approved to run production data.

The problem: A federal institution's data — including PII — lives under the strictest security and compliance requirements in the country, on legacy on-premise systems. Cloud was considered out of reach for production data.
What we built: A greenfield, cloud-first platform designed to pass where none had before: hybrid on-premise + cloud architecture, confidential VMs with strict networking, everything as infrastructure-as-code — covering ingestion, orchestration, lakehouse, visualization and AI/ML, with the governance layer built in and a developer experience teams actually enjoy.
Where it landed: First of its kind approved for production among the federal instances. We designed it in a Lead Data Architect role and lead it today as Head of Data Platform — while the legacy-to-modern-stack migration it enables is underway.

AI Strategy

Building AI that understood customers as people — not segments — in a fintech product we built and ran.

The problem: Financial institutions knew their customers' transactions, not their people. CRM data said what a customer did — never who they were, or which salesperson could actually reach them.
What we built: A cloud-first B2B SaaS: an AI engine performing automated persona building — interest profiles grounded in psychology research — surfacing insights for sales teams and automatically matching every customer with the right salesperson. We built the architecture, backend and AI/ML models ourselves.
Where it landed: The skills it proved: taking AI from research literature to a revenue-facing production system — the standard we still hold AI advice to.

Security Strategy

Making 'secure enough for production' achievable where cloud was considered impossible.

The problem: The most valuable data is the most restricted: PII under government-grade security requirements, where a single gap keeps a platform out of production forever.
What we built: Security architecture as a first-class part of platform design, not an afterthought: confidential VMs, strict network isolation, IAM and secrets management, auditability by design — everything as infrastructure-as-code, shaped alongside government security-architecture communities.
Where it landed: The skills it proves: security strategy that unlocks approval instead of blocking it — the same discipline behind our security reviews of data environments and AI systems.

Product Engineering

Designing, shipping and operating live products — enterprise platforms and consumer apps alike.

The problem: Products fail less on code than on everything around it: operations, discovery, partnerships, and the discipline of running production software day after day.
What we built: Two live products built in-house: an enterprise platform with time-series analytics at its core (Go, Python, React) that gives organizations visibility over their software landscape — and a full-featured consumer mobile app with discovery, events and tournaments, partner venues, and the admin portal that keeps a live community running.
Where it landed: The skills it proves: platform architecture, product thinking, mobile engineering and the operational discipline of production software we answer for ourselves.

Client engagements are confidential by default — we describe the work, never the customer.