Engineer. Founder. Builder.
Turning complex AI
into systems that work.
I build production AI, lead technical teams, and share what I learn along the way. Founder & CTO of AlexandrIA. Based in Paris.
01 / In practice
A research copilot that connects knowledge, sources, and ideas. Built end to end, from retrieval and multi-agent orchestration to evaluation and EU-sovereign hosting.
Working with teams
As a Fractional Head of AI, I help post-seed product teams make the right technical decisions and ship reliable systems.
Roadmap, architecture, evaluation, and delivery. Typically one to two days a week.
About working together02 / Field notes
Practical feedback loops for AI-assisted development, from Claude Builders Night in Paris.
Hardware, datasets, models, and the dependencies behind the stack. A course at Centrale Nantes.
VAEs, diffusion models, GANs, and transformers: the ideas that connect them and the tradeoffs that matter.
03 / In conversation
The podcast · 24 episodes
Honest conversations with engineers and founders about the decisions, failures, and lessons behind their work.
Explore the archiveSpeaking & teaching
From TensorFlow World to Sorbonne Université and Centrale Nantes. Talks and courses on AI systems, model training, and engineering choices.
Talks, slides & replaysA little context
I’m François, a founder and engineer based in Paris. I graduated from École Centrale de Nantes with an MSc in Signal Processing & Statistics.
I spend my time building AlexandrIA, helping teams bring AI into production, and teaching LLM pretraining and post-training at SCAI (Sorbonne Université) and Centrale Nantes.
This is where I keep the useful things: technical notes, talks, and conversations with people whose work I find interesting.
Find me on GitHubLet’s compare notes
For an AI challenge, a speaking invitation, or a good conversation.
Get in touch