Groovematter exists because we got tired of being recommended music we'd already heard, from artists we already knew, in genres we were already deep in.
Every recommendation engine on the market works roughly the same way: it looks at what millions of other people listen to and assumes you want the overlap. For casual listening, that's fine. For anyone who actually cares about music — DJs, diggers, collectors, producers — it's useless.
We built something different. An engine that learns your specific taste — not a genre, not a demographic, not an average — and finds music that belongs to you. Music that no playlist, no chart, and no “you might also like” would ever surface.
We use deep learning audio analysis — similar to how Shazam identifies songs, but pointed in a different direction. Instead of asking “what is this track?” we ask “what does this track sound like, and what else in the world sounds like it in ways no tag can describe?”
Combine that with the largest dataset of real music curation decisions we could assemble, and you get recommendations that understand context, texture, and energy — not just genre and BPM.
Interactive 3D taste field — navigate your sound in space
Weekly personalised digs — fresh finds for each of your fields
Set builder — energy arc, harmonic mixing, BPM flow, automated
Deeper audio analysis — mood, texture, rhythmic complexity
650K+
100K+
5.6M
16K+