A personal music intelligence layer- built on top of your listening history.
Lyra is a standalone application that sits alongside platforms like Spotify and YouTube Music.
It does not replace playback.
Instead, Lyra transforms your fragmented listening history into a clean, structured, and controllable system, allowing you to generate playlists based on your actual taste.
Your music apps know what you listen to. Lyra lets you use that knowledge intentionally.
-
🔌 Connects to your music platforms
-
📥 Ingests your listening history
-
🧹 Cleans and structures your data
-
🎛️ Lets you filter by:
- language
- energy
- recency
-
📜 Generates dynamic playlists
-
🔗 Opens tracks back in the original app
- ❌ Not a music player
- ❌ Not a recommendation engine (yet)
- ❌ Not trying to replace Spotify
- Flutter
- Python (FastAPI)
- Go (orchestration, future services)
- PostgreSQL
- Podman (preferred)
- Docker (fallback)
- User connects Spotify
- Lyra fetches listening history
- Data is normalized + stored
- Enrichment adds metadata
- User applies filters
- Playlist is generated
- Tracks open in Spotify
GET /auth/loginGET /auth/callback
GET /sync/recentGET /sync/top
POST /playlist/generate
POST /presets/saveGET /presets
- Spotify-only integration
- Basic language detection
- No advanced intelligence
- No cross-platform deduplication
- mood classification
- improved tagging
- embeddings
- similarity search
- dynamic sequencing
- context-aware playlists
- full taste identity system
- adaptive playback intelligence
understanding your relationship with the music you already love.
MIT (or TBD)