Keeping a high-traffic product alive
L'Étudiant
When tens of thousands of people use a product, the smallest regression shows instantly. That's the ground I worked on at L'Étudiant, the platform connecting future students with enrolled ones for honest reviews of schools.
01The problem
The product ran on an old, complex codebase, loaded with tech debt. Adding features to it without breaking anything is the trickiest exercise there is.
02What I shipped
I stepped into the existing code and shipped: back-end and API in Ruby on Rails, web in React / Next, mobile in React Native. Review system, guidance paths, matching. Every time working with the legacy, never around it.
03Key features
- Full-stack development (Rails, React / Next)
- Mobile app (React Native)
- Review system & guidance paths
- Changes on legacy, no regressions
Result : Features shipped and maintained for an audience of tens of thousands, on a complex codebase, without ever breaking prod. Proof that a fragile codebase can be trusted to me, no hesitation.