Advisory
Strategic advice from someone who still writes the code.
How I work
Engagement model
Most of my advisory engagements run for one quarter. Long enough to go deep, short enough to leave decisions with the client. Within that window, the format varies — sometimes it's a single intensive review and a short report; sometimes it's a weekly cadence and a seat in the team's architecture meetings.
I do not run multi-month transformation projects. I am not building a consultancy that scales by adding analysts. The engagement is with me, not with a team you've never met.
I work in English, Dutch (NL/BE), or French as needed.
What I'm useful for
Three kinds of engagement
The pre-spend second opinion. Teams that are about to commit serious budget to an AI initiative and want a sharp review from someone who has shipped this work themselves. I audit the architecture, the build-vs-buy assumptions, and the procurement framing. Deliverable is usually a short written report, two or three working sessions, and a weekly call across one quarter. Most of the value lands in the first three weeks; the rest is following through on the decisions made.
The unit-economics review. Teams that have AI in production but aren't happy with what it costs to run. I help them understand where their money is going — usage breakdown, model choice, prompt structure, retrieval cost — and decide what to keep, replace, or remove. This is the most common engagement, and it almost always ends with a smaller AI footprint than where it started.
Fractional Head of AI. Teams that need a senior AI voice in their leadership cadence but don't want to hire one yet. I sit in your weekly leadership meeting, your architecture reviews, and your vendor calls. Up to one day a week, capped at a quarter to start. Renewable, but only if it still makes sense for both sides.
When I'm not the right fit
Self-qualifying clarity
I don't write white papers, run AI literacy workshops, or build slide decks for boards I'm not part of. I'm not a fit for organizations that want validation rather than scrutiny. And I'm not interested in engagements where the answer is already decided.
If your AI work is governance-heavy with no technical work yet, you want a different advisor.
A note on the operating role
Why I keep the day job
I'm Partner and Head of Data & IT at Dewaele. That's a real job — hundreds of users, customer-facing systems, my pager goes off when something breaks. Advisory clients sometimes ask whether this is a conflict. The honest answer is that it's the opposite: the day job is what makes the advice worth paying for. Strategy I publish or recommend has to survive its own deployment, in my own shop, before it shows up in your meeting.
Sinax engagements are capped to keep this true. I take on no more than four active advisory clients at a time.