About me

The short version

I help founders turn business direction into AI systems their teams can use. That means moving between the founder, developers, marketing, operations, product, and the technical build without losing the goal.

Ecommerce is the business I know most deeply, and marketing is my strongest area. I spent years doing copywriting, market research, automation, and growth strategy for 7 and 8 figure brands. Now I use that experience to build company brains and autonomous systems around real work.

I make YouTube videos about what I am learning (34K+ subscribers). I live in the French countryside with my wife Sally, our daughters Léa and Julia, our son Milo, and our dog.

Why systems

I'm a tinkerer. I can spend three hours trying to make something work just to understand how it works. When AI agents became good enough to do real tasks, I couldn't stop building them.

What surprised me was how much understanding the business mattered. I can listen to a founder, clarify what should change, and follow that goal through the people, tools, and decisions involved. That lets me judge whether a system is helping instead of stopping at whether the technology runs.

I eventually found a name for the role: fractional forward deployed engineer. In plain English, I work inside a company, close to the team and the problem, and turn what I learn into systems people can use. The intersection of business knowledge and system building is where I do my best work.

What I care about

Freedom. Not in a "laptop on a beach" way, but in a "nobody tells me what to work on" way. My biggest value is autonomy, and it's not negotiable.

Deep work. I need three-hour blocks of uninterrupted focus to do my best thinking. That's when the real stuff happens.

Learning by doing. I'd rather build something wrong and fix it than spend a week planning the perfect approach. You learn more from the building.

The long version

No degree. No formal education in anything I do today. I discovered I actually love learning around age 30, which is late, but apparently not too late.

I joined the French army at 18. Infantry. Nine years, multiple deployments. While I was serving, I studied on my own to get construction qualifications, going from no credentials to certified in a few years of evening study. That stubbornness turned out to be useful later.

After the army I ran construction teams for five years. Good work. I learned how to lead people, deliver projects, and deal with clients who change their mind every Tuesday. But I wanted something different.

So I tried building things online. A dropshipping store that took off fast, then fell apart just as fast. Two years of experiments that mostly didn't work. Then a eucalyptus bedsheet brand that I still think was a great product. Customers loved it. I still use those sheets five years later. The economics were hard, €30K minimum order for one color and six months to produce. But that is not why I stopped.

Through all of that, I was doing the marketing myself. The copy, the ads, the funnels, the research. An agency founder noticed what I was doing during the bedsheet pre-order campaign and invited me to help with their clients. Six months later I had six clients of my own.

That's how consulting started. Not from a plan, but from people seeing the work and asking for more of it.

While the pre-orders were being produced, I had started consulting to make a living. Eventually I had a real choice: keep pursuing my own ecommerce brand, or continue consulting, where I already had clients.

I chose consulting. My mind likes variety. I love learning, moving between different problems, and going deep enough to make myself useful. Consulting let me do that. For years I worked with long-term clients and helped build their brands.

Gradually, more of my work became automation. First with no-code tools, then with systems I built for ecommerce brands before AI was part of the picture.

When GPT-3 arrived, before ChatGPT, I used it immediately. I realized I needed to spend my time understanding how AI worked and what it could do. So I did, every day.

When agents arrived, the pieces finally connected. For the first time, I had systems that could deliver real work. I understood how they worked, and I understood the business work they needed to do.

I dropped everything else and focused on understanding how agents could be deployed inside companies. That is what I work on every day now.