Description:
Spotify Playlist Curation Tool that transforms how users experience their music libraries. Traditional playlist organization methods based on simple genres or artist names don't always capture the nuanced vibes listeners seek. The tool analyzes users' existing playlists, generates dynamic categories based on song metadata and genres, and curates new, personalized playlists tailored to diverse listening preferences by leveraging LLM capabilities.
- Python version 3.10+
- NodeJs version 22
Install for Macbooks
cd server
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
cd ..
cd web
npm ci
Install for Windows
cd server
python -m venv venv
Set-ExecutionPolicy Unrestricted -Scope Process (This would allow running virtualenv in the current PowerShell session.)
venv\Scripts\activate
pip install -r requirements.txt
cd ..
cd web
npm ci
Create .env file with the following content:
CLIENT_ID=...
CLIENT_SECRET=...
REDIRECT_URI=http://localhost:3000
OPENAI_API_KEY=...
Open two terminals or powershells
Terminal 1:
python app.py
Terminal 2:
npm start
To run the unit tests
# navigate to the test folder
cd tests/unit/
# run the test
pytest test_unit.py
- Click on "Login to Spotify" in the center of the page and log in to your Spotify account.
- Choose a playlist you want to classify. The playlist name and the number of songs will be displayed on the left and right sides.
- Select the number of categories you want to classify, which should not exceed the number of songs in the playlist. Click on the "Generate" button to generate category names and descriptions based on your selection.
- Wait for the AI classification to complete. You can track the progress of the classification from the progress bar at the top.