Un modèle précis à 95 % dans un notebook et qui n'atteint jamais la production ne vaut rien. Pour les équipes de Quartier Latin, je construis le pipeline complet — entraînement, déploiement, surveillance — pour que le ML soit réellement livré.
Get in touch
Tell me about your goals — I'll reply within 24 hrs.
What's your biggest challenge as a developer right now?
What's your biggest challenge as a developer right now?
Which framework do you prefer for 2025?
Which framework do you prefer for 2025?
The Dark Patterns We Still See Everywhere (And How to Avoid Them)
Dark patterns are UX designs that trick users into doing things they didn't intend to. Th…
AI won't replace you. A person using AI better than you wil…
AI won't replace you. A person using AI better than you will. The skill shift happening r…
How do you find your first 100 customers without a marketing budget?
Bootstrapping a B2B SaaS. No budget for ads yet — need to get traction organically. I've …
Hot take: most landing pages have too many words and not en…
Hot take: most landing pages have too many words and not enough proof. Your visitor doesn…
Why I Switched From Vercel to a Self-Hosted Setup (and Why You Might Not Want To)
Six months ago I moved my production apps off Vercel and onto a self-managed VPS. Here's …
Reminder: the best portfolio is shipped work. Not a beautif…
Reminder: the best portfolio is shipped work. Not a beautiful Figma file no one has ever …