Personal music knowledge base
A private library of enriched tracks — searchable, playable, explorable. Built with a full ingestion pipeline and AI-assisted metadata.
Catalog snapshot — live totals load when the database is connected
For everyone
Your music, organized and enriched. Sign in to browse, stream, and explore the full library — Settings is for admin accounts only.
Volumetry
Representative catalog figures — connect the database for live totals.
Top moods
Top decades
| Computer | Tracks | GB |
|---|---|---|
| macbook-2016 | 2,100 | 28 |
| macbook-2023 | 1,900 | 24 |
| other | 1,557 | 20 |
ML clusters by lens
Playlists
How it works
From raw MP3s on S3 to a searchable knowledge base. Runs on your Mac; the web app reads precomputed data.
Tools
S3 scan, CSV, iTunes XML
Output
Recordings, artists, playlists, enrichment jobs
LLM
No — deterministic
Engineering
Four app modules over a Postgres knowledge base. Pipeline runs locally; Heroku serves the UI.
┌─────────────┐ ┌──────────────┐ ┌─────────────┐
│ S3 (audio) │────▶│ Mac pipeline │────▶│ Postgres │
└─────────────┘ │ ingest→postcluster│ │ knowledge │
└──────────────┘ └──────┬──────┘
│
┌─────────────┐ ┌──────────────┐ │
│ OpenAI │◀───│ Next.js 15 │◀───────────┘
│ facts/curate/identify│ │ Heroku (web) │
└─────────────┘ └──────────────┘AI layer
LLM-assisted enrichment with citations and cached responses. Operational cost details stay in admin Settings.
Per-track genre, mood, decade, and description from MusicBrainz, Wikipedia, and LLM synthesis — with source URLs.
Unlabeled or filename-only tracks get Whisper transcription plus web-search identity with confidence scores.
SHA256 exact-match and acoustic fingerprint candidates — reviewed manually before any merge.
Collections named and described by a reasoning model; post-cluster LLM proposes merges and completions.
Natural-language queries translated to structured filters for semantic search.
LLM responses cached by prompt hash. Pipeline runs track usage for operator review in Settings.
LLM cache
—
Cache stats appear when the database is connected.
Usage analytics
Models in use
Status
A living project. Core pipeline and library are built; usage analytics and sharing are next.
7 of 10 milestones shipped
Cognito sign-in; per-account playlists and listening
Tracks, artists, collections unified
MusicBrainz + LLM with citations
Four lenses, fused coordinates
Fingerprint + transcription fallback
Public overview with live stats when database is connected
Plays, heard/completed, listening dashboard
Interactive playlist candidate review
Curated read-only views
Offline queue, installable
Personal project. Work in progress. Not a public streaming service.
© 2026 Zics Vault · Built with care in Paris