← All work

Side project · Full-stack

Turning listening history into better recommendations

Timeline
2026
Team
Solo
My role
Design, backend, data, and deployment
Skills
Python, FastAPI, Neon Postgres, Spotify OAuth

Build an analytics platform on top of my own Spotify data: sign in with Spotify, store listening history, and use it to power a recommendation engine.

  1. 01

    Real sign-in, real data

    Users authenticate with Spotify OAuth, and the FastAPI backend pulls their listening data into a Neon serverless Postgres database.

  2. 02

    A recommendation engine on top

    With the history stored, a recommendation engine suggests music based on what each user actually listens to.

I'd like to add a way to measure whether the recommendations are good, beyond whether they feel right.