Results-driven AI, grounded in real operating experience.
Over two decades of leadership roles in marketing and e-commerce, including managing one of France's top e-commerce operations at Disney. Now I build custom AI systems for businesses, and get their teams good at using them.
Most AI projects fail for the same reason.
It is rarely the tools. It is the lack of business context and method behind them. Ask a generic model a real business question and you get a generic answer: plausible, confident, and too shallow to act on. That holds whether you're reading your analytics, making sense of customer feedback, or producing content at scale. The value is in the engineering: systems built around how a specific business actually works.
What custom actually looks like
Most AI work stops at a chatbot or a prompt template. These are full systems, engineered around a specific business problem and built to run in production.
A content engine with AI search built in
For a European technology company, I built a content generation system that lifted both the volume and the quality of their marketing output: blog posts, social, the lot. Phase two went further, embedding a full layer of AI search optimization (GEO) into the engine, so every piece is engineered to be found and cited by AI search, not just published. A custom interface sitting on a structured database, with Claude Opus under the hood.
A lead system that writes itself, per prospect
For a client fielding inbound that needed sorting, I built a lead generation system that turns a single visitor input into a personalized report. The visitor shares one detail about themselves, the system pulls live data from a public source in real time, generates an editorialized report tailored to them on the fly, and delivers it with conversion copy tuned to move them forward. Useful enough that people want it, specific enough that it converts.
A front door that sorts your leads for you
For a client fielding mixed-quality inbound, I built a qualification system that reads each visitor's answers and routes them down the right path automatically: a reality-check guide for researchers, next-step resources for warm leads, a fast-track guide and booking link for the ones ready to move, priority escalation for enterprise. The routing is the easy part. The work is in the system generating the right tailored material for each segment, so every lead gets a genuinely useful response and you only spend time on the ones worth it.
A career building businesses, sharpened by AI.
I ran DTC e-commerce for Disney across France and Benelux, scaling it into the top 4% of French e-commerce. A decade-plus in senior strategy and marketing roles before that, and three startups cofounded along the way. I've been hooked on generative AI since the early days, not as a vendor chasing a trend, but because I could see what it would do for businesses like the ones I'd spent my career running. That's the lens I build from: real excitement about the technology, disciplined by years of being accountable for results.
Three ways in
Custom AI builds
Systems engineered around your business, like the ones above. Scoped to a real problem, built to run.
AI enablement and training
Full-day, hands-on sessions that take marketing teams across the tools that matter: content, image and video generation, AI for analytics, AI for customer research. Not a tour of features, a working command of the kit they'll use every week.
Embedded consulting
Sitting in with your team to find where the real blockages are, then designing the AI-enabled workflows that make them faster and better at the work they already do.
Let's talk
If you have a problem worth solving with AI, or a team that needs to get sharper with it, get in touch.
No sales funnel, no account manager. Every note reaches me directly, and I reply within a day or two.