Inspiration

During major tournaments, fans constantly juggle logistics and decisions: where to stay, what to eat, what to do between matches, and how to evaluate betting opportunities responsibly. We wanted one assistant that could turn that chaos into a clear plan.

What it does

World Cup Intelligence is a planning and decision-support app for tournament travel and match days. It combines:

  • Stay/food/activity planning
  • Match context (form, news, weather, availability)
  • Odds analysis with model vs market comparison
  • Risk-aware recommendations and bankroll simulation (paper trading by default)

How we built it

We built a multi-service system with:

  • FastAPI backend + polling worker
  • SQLAlchemy/Alembic + Postgres
  • Streamlit dashboard
  • Modular services for odds ingestion, modeling, risk controls, recommendations, and audit logs
  • Docker Compose for local orchestration
  • Automated tests for calculations, risk logic, and model behavior

Challenges we ran into

  • Balancing scope between travel-planning features and betting analytics
  • Managing context/token limits while iterating quickly with AI-assisted coding
  • Integrating multiple subsystems into one coherent user flow under hackathon time pressure

Accomplishments that we're proud of

  • Built a real end-to-end product, not just a prototype screen
  • Unified several independent components into one storyline and workflow
  • Added practical safeguards like simulation mode, risk controls, and auditability

What we learned

  • A clear architecture matters more than raw code volume in a hackathon
  • AI agents are powerful for speed, but require strong structure and validation
  • The best demos solve a concrete user problem from start to finish

What's next for World Cup Intelligence

  • Expand coverage to more tournaments and leagues
  • Improve personalization for itinerary + betting/risk profiles
  • Add stronger real-time data integrations and smarter recommendation ranking
  • Pilot with early users, then scale through partnerships and product iteration

Built With

Share this project:

Updates