How Identify Genre classifies music — the data pipeline

Pipeline overview

Identify Genre is a server-side report builder wrapped in a simple paste-a-URL UI.

Spotify URL → parse artist ID → identity fetch → enrichment cascade → report JSON → UI tabs / export.

You stay on one page; the heavy work happens in `/api/analyse`.

Stage 1 — Resolve

Accept artist or track URLs. Track URLs resolve to the primary artist so marketers can paste whatever link they have open.

Stage 2 — Identity

Fetch Spotify artist identity: name, images, external URLs. This anchors the report even when genre fields are empty.

Stage 3 — Enrichment cascade

Enrichment layers add genres, similar artists, and market hints:

• Last.fm tags and similar artists
• MusicBrainz structured metadata
• Every Noise / genre-map context
• Fallback inference from similar-artist tag overlap when primary sources are thin

Caching (for example Redis) may store recent reports to cut repeat API latency.

Stage 4 — Product surface

The UI splits the datasheet into tabs: Overview, Genre, Similar Artists, Playlist Targets, Listenership Map (and related targeting views). Exports serialise the same structured report to PDF or CSV.

Why this architecture

• No login — lower friction for one-off research
• Cascade — resilient when a single API is sparse
• Export — the report has to leave the browser for real workflows

Related reading

• Spotify vs MusicBrainz vs Every Noise
• Genre targeting for music marketing
• How it works

Run a live report from the analyser.