Business first, code second
The best technical decision I ever make on a project is usually understanding the domain properly. DDD isn't a diagram style. It's a way of listening.
I'm Nizar. I grew up in Tunisia, studied computer science in Sousse, and now live in Paris, where I write software for a living and make music for no reason at all.
I got into engineering because I liked taking things apart. I stayed because of the opposite: the feeling when a dozen moving pieces finally click into one thing that works. That feeling hasn't worn off in six years.
Along the way I've built delivery platforms that route trucks across cities, customer spaces for a national postal service, salary tooling for an international organisation, and cashflow forecasting for small businesses. Very different rooms. In each one, my job was the same: understand what people are really asking for, then build it so well that they stop thinking about it.
I'd rather ship something simple that people trust than something clever that only I understand.
I'm the kind of engineer who asks "why" a few more times than is comfortable, who actually enjoys a good code review, and who will happily own a feature from the first whiteboard sketch to the 2am incident. I've been the technical referent, the product owner, and the intern. I learned something in every seat.
The best technical decision I ever make on a project is usually understanding the domain properly. DDD isn't a diagram style. It's a way of listening.
Hexagonal, CQRS, event sourcing. I use them where they earn their keep, so the codebase stays easy to change two years in, when nobody remembers why anything is the way it is.
Data model, API, UI, pipeline, dashboards, incident. If I built it, I want to know how it behaves in production. Observability is part of the feature.
I've set up and led code review practices on several teams. Done well, a review isn't a gate. It's two people making each other a bit better.
I work with AI coding agents every day, and I ship real features with them. What makes it work is spec-driven development: clear intent written down first, then let the agents run.
Shared libraries, project templates, CI that's actually fast. A team that isn't fighting its tools ships better software and goes home on time.
Where I'm most at home.
How I like to shape systems.
Getting information from A to B, intact.
Because "works on my machine" isn't a deployment strategy.
Knowing it works, and knowing when it stops.
The newest tool in the box, and I use it daily.
At engineering school I spent as much time in the Junior Entreprise as in lecture halls. Running a student-led company taught me things no course did: how to sell an idea, lead a team, and handle a client who changed their mind.
I'm still part of its alumni foundation today. Giving back to the place that gave me my start feels like the right thing to do.