DEV Community

Cover image for I built a massive sports data engine from my truck cabin (Supabase + Monte Carlo)
Paul Iroveanu
Paul Iroveanu

Posted on

I built a massive sports data engine from my truck cabin (Supabase + Monte Carlo)

Hi DEV community! šŸ‘‹

I’m Paul, an international truck driver by day, and a solo maker by night. I just launched Statlyzer, a transparent football data analytics engine, and I literally built the entire backend architecture directly from the cabin of my 40-ton truck during my mandatory highway rest breaks across Europe. šŸš›šŸ’»

šŸ› ļø The Tech Stack & Architecture

I wanted to build something extremely robust that doesn't lag on mobile devices despite running heavy math.

  • Database: PostgreSQL hosted on Supabase. (Migrated away from Firebase to handle complex relational queries cheaper and faster).
  • The Engine: It runs high-iteration Monte Carlo simulations, Poisson distribution, and Dixon-Coles models.
  • The Output: Raw probability percentages for variables like 1X2, Over/Unders, corners, and cards across global fixtures.

šŸ” The "Past Analyst" Feature

The sports prediction market is full of fake gurus hiding their losses. I built Statlyzer to be 100% transparent. The app automatically logs and publicly archives every single AI prediction from yesterday against the real-world results. If the mathematical model misses, it stays on the dashboard for everyone to see. No deleted data.

šŸ’” The Model

It's a freemium app. Every user gets 2 Free Credits daily to fully analyze any 2 matches. For power users, I integrated Stripe/Paddle for subscriptions.

I would absolutely love for you guys to test the web demo and try to break it. Give me your brutal feedback on the database responsiveness and the UI!

šŸ”— Web Demo: [app.statlyzer-studio.com]https://app.statlyzer-studio.com/
šŸ”— Official Site: https//statlyzer-studio.com

Thanks for reading, and I'll be answering questions in the comments as soon as I park the truck!

Top comments (0)