← Back to Home Projects

MovieMatch

A Streamlit-based movie recommendation platform that uses item-based collaborative filtering to generate personalized suggestions, plus a lightweight achievement system and social discovery to make the experience engaging.

Python Streamlit scikit-learn SQLite Collaborative Filtering TMDb API

What I built

MovieMatch brings together a recommendation engine, user accounts, and a clean UI to help users discover movies faster — with explainable logic and practical product features.

  • Personalized recommendations using cosine similarity on an item–item matrix.
  • User accounts with hashed passwords and stored rating history.
  • Badges & gamification to reward engagement and repeat usage.
  • Social discovery by matching users with similar rating patterns.
  • Poster + metadata enrichment via TMDb API with caching for faster loads.