Spotify Nebula 🪐

Visualize and discover more of what you love listening to.

     

🛠️ Features

• Implemented relational database support for PostgreSQL/SQLite3 using SQLAlchemy to safely store user authentication tokens, spotify access tokens, and expiry timestamps.
• Utilized a Python/FastAPI backend for external API calls (Spotify Web API, SoundNet API) and internal endpoints.
• Pipelined dense quantitatve audio data through dimensionality reduction and clustering algorithms (t-SNE and DBSCAN) using numpy and scikit-learn.
• Created an interactive dashboard using ReactJS to visualize a user's Spotify taste nebula.

🚀 Tech Stack

• Languages: Python, Javascript, HTML, CSS
• Database: SQLite3(inital development), PostgreSQL(migrated to later)
• Frameworks/Libraries: FastAPI, ReactJS, SQLAlchemy, Pydantic, Scikit-learn
• APIs: Spotify Web API, SoundNet Audio Feature API

Demo Screenshots

• Final dashboard built with ThreeJS and React

Development Screenshots

• Initial visualization with Plotly

• Link to demo: https://drive.google.com/file/d/11vaPjYpVOHIJO-LCL2yXALcr2TxTJ_oA/view

⏰ WIP

• Optimized clustering algorithim of choice
• Deprecated Spotify audio features endpoint workaround

Collaborators

• Evan Sun
• Ryan Liu