I started programming to understand how games and software worked on the inside. That led me to web development, systems, and then artificial intelligence.
Today I build complete applications, from frontend to backend, with a specialisation in computer vision: I train the models and build the products that use them.
I worked as a freelancer for FaceIQ Labs, I'm a co-founder of DiscordGate, and after two years of a computer science degree in Rennes, I joined 42 Paris to go further in systems programming.
FaceIQ Labs is a US startup that analyses facial proportions from photos. Its platform reports more than 1.3 million users and 1.4 million analyses.
I worked there as a freelancer, around 30 hours a week. I rebuilt the facial harmony report, added sharing for reports and results, and fixed security vulnerabilities and bugs in production.
Next.js
TypeScript
Tailwind CSS
Prisma
PostgreSQL
Selected projects
Products I've worked on, alone or with a team.
private codeIn development · Pre-release
Glow Up: My Beauty
Mobile app · Independent project
A mobile app that brings together colour analysis, hairstyle simulation, visual skin analysis and skincare routines.
I designed the experience, built the mobile interfaces and integrated the analysis and image generation features.
A facial analysis engine I built alone, end to end. At its core, a facial landmark model trained from scratch with PyTorch; its predictions are used to compute facial measurements and proportions.
I handled the whole chain: data, training, evaluation, inference, API and web interface.
For training, about 30,000 images were pre-annotated by MediaPipe, then 400 corrected by hand: 75 times cheaper than fully manual annotation.
A marketplace for renting sports cars, supercars and hypercars in the Île‑de‑France region, between professional rental companies and vetted drivers.
As one of three partners, I designed and built the public site and the pre-registration flows for rental companies and future customers.
Next step: the booking platform, with vehicle search, deposit payments and handover inspections.
Public site live, waiting list open.
Next.js
TypeScript
Supabase
Drizzle
PostgreSQL
private codePersonal projects · Technical archive
Reverse engineering & experiments
Systems programming · Networking · Computer Vision
Before building products, I spent a lot of time taking video games and software apart to understand how they work inside.
Rebuilding a game server from its observable behaviour, network analysis tools, reading process memory, prototypes combining visual detection with Arduino devices.
That is where I learned low-level programming, and how to understand a system with no documentation at all.
BSc Computer Science · 2024–2026 · First two years completed
Two years of fundamentals: algorithms, data structures, object-oriented programming, systems and networks, databases, functional programming and statistics.
Languages studied : Java · Python · C · Haskell · R · SQL · JavaScript