Education
CentraleSupélec, engineering and entrepreneurship
An engineering education, an entrepreneurship medal and an early desire to build products rather than stop at the idea.
About
Turning knowledge into systems.
An engineer, entrepreneur, AI speaker and trainer, I work at the intersection of technology, strategy and hands-on delivery to help organisations make realistic use of generative AI.
I came to AI through engineering, entrepreneurship and technology commercialisation. My starting point is rigorous knowledge, sometimes from research, and the challenge of making it useful in a real organisation.
I founded my first company while studying at CentraleSupélec, developing medical devices subject to ANSM requirements, France's medicines and health products regulator. Since then, I have worked across different technical and entrepreneurial settings, with the same guiding principle: understand a technology in depth, then put it to the test in practice.
What interests me about generative AI goes beyond model performance. It is what businesses do with it: what they automate, what they keep under control, which skills they develop, which rules they set and how much value they actually create.
What drives me
The vertigo of machine cognition.
When the first large language models began showing a real command of human language, I felt a very particular sense of vertigo. I knew this would happen, but I did not expect to experience it so soon. There was a tipping point, a symmetry-breaking moment when everything began to accelerate.
For me, we need to return to a simple idea: these systems are a form of machine cognition. In a largely service-based economy, where so much value comes from human cognition, this is far more than a technical detail. It profoundly changes what organisations can automate, augment, decide and produce.
That is also why magical promises and lazy analysis both frustrate me. AI deserves more than vague enthusiasm or reflexive panic. It calls for precision, realism and the ability to connect ideas with practical uses.
The most fruitful ideas often emerge when two worlds meet. I like this phrase: mathematics is the art of giving the same name to different things.
Background
First company
Medical devices: an early reality check
Building medical devices teaches you early that a technology's value goes beyond its demonstration: it must be robust, documented, responsible and usable.
Tech career
Founder, CTO, complex products
Software, finance, regulated products and sensitive business contexts taught me to assess innovation through its architecture, data, team, market and accountability constraints.
Generative AI
From fascination to action
GPT-2, then GPT-3.5 and ChatGPT, made a profound shift visible: some repetitive cognitive work can now be automated. The question goes beyond understanding AI to learning how to integrate it properly.
Beliefs
Connect disciplines
I have always been fascinated by people who can draw on several fields of knowledge. The most fruitful ideas often emerge where disciplines meet.
Take a realistic view of AI
AI sparks as many fantasies as fears. My role is to help distinguish what is useful, what is risky, what is immature and what is truly worth building.
Put technology to work
Creating value means taking the best of research and technology and making it useful to as many people as possible.
What I bring
A technical perspective without unnecessary jargon, an entrepreneurial mindset and a persistent focus on turning AI's promises into real uses, clear decisions and systems that last.