3 min read
What I Learned Building AI Products at 25

I’m 25, based in Paris, and I spend most of my free time building things nobody asked for.

By day, I’m a software engineer at Advanced Schema consulting for LVMH — ML pipelines, microservices, the usual. By night (and weekends), I ship AI apps. An AI wardrobe that extracts clothes from outfit photos. A CV tailor that runs entirely in your browser. A chart generator that won a Plotly challenge and somehow hit 25k downloads.

None of these make money. But they taught me more than any job could.

The builder’s trap

Here’s what nobody tells you about being a “builder”: shipping is easy, distributing is hard.

I can go from idea to deployed product in a weekend. I’ve done it multiple times. The problem is that I do it too much. I start something, ship it, feel the dopamine hit, then move on to the next shiny thing.

My GitHub has 27 repos. Most of them are noise. The few that matter — dripin, CVPilot, ChartGPT — could be much bigger if I focused on one instead of constantly chasing the next idea.

What actually matters for your career

After spending way too long thinking about this, here’s what I’ve landed on:

  1. A name on your CV beats a project on your GitHub. Sad but true. Having “Datadog” or “Mistral” on your resume opens doors that 33 GitHub stars never will.

  2. Generalists are useful, specialists are hireable. I can write Python, TypeScript, Go, Java, React Native, and deploy on GCP, AWS, and a Raspberry Pi in my closet. That makes me useful on any team. But it doesn’t make me the obvious choice for any specific role.

  3. Your salary is a function of your negotiation, not your skill. I’ve met people who are technically worse than me making 2x my salary. The difference? They know how to sell themselves. They never give their current salary. They always have multiple offers.

  4. Ship one thing well > ship five things halfway. This is the hardest lesson for me. I’m still learning it.

What’s next

I’m focusing. For real this time.

  • Getting dripin on TestFlight and into real users’ hands
  • Looking for my next role in AI engineering (if you’re hiring, let’s talk)
  • Writing here, occasionally, about what I learn along the way

If you’re a young engineer in the same boat — building a lot, distributing nothing, wondering if it matters — just know you’re not alone. The projects matter. They just matter more when you can tell the story around them.