DEV Community

Yasser Mbarek
Yasser Mbarek

Posted on

How I built a Moroccan Legal AI with $0 to beat a $200k Startup

The Backstory I had this idea a while ago an AI that explains Moroccan Law to normal people. I started coding it in my bedroom, taking my time, trying to make it good.
Then, the worst thing happened.
While I was still working on mine, another startup launched the exact same concept. They were faster than me. A few months later, they were on national TV, raising $200k (approx. 2 million MAD) from investors.
I Almost Quit I thought it was over. They had the money, the fame, and the "First Mover" advantage. I’m just a 19-year-old student with a $100 budget.
But then I actually used their app.

  • It felt heavy and slow.
  • The UI was cluttered.
  • The AI actual answers where harder to read than the original law scripture.
    I checked their Google PageSpeed score. It was 28.
    I realized that "First" doesn't mean "Best." They might have the funding, but that made them slow and corporate. I had nothing to lose, so I could focus purely on the code.
    I decided to finish my project anyway, but with one rule that it has to be faster than theirs. cause theirs take 2-5 min to answer but mine ~10 sec.

  • Frontend: Next.js

  • Data: I wrote a Python script to scrape the official government law texts myself (Official Bulletin & SGG).

  • Performance: I stripped away all the bloat. No tracking scripts, no long logins.

I finally launched 9anon AI.

I'm sharing this because I almost stopped working on my dream project because I saw someone else "win." But money can't buy a clean codebase.
If you are a student and you see a big company doing what you want to do don't stop. You can probably build it faster than them.

Check it out here: https://9anonai.com
(And yes, the data scraper is open source, check it out in https://github.com/berserker-glitch/moroccan-law-data)

Top comments (0)