Music Recommendation Audiogen

Audiogen is a semantic audio technology enabling the recommendation of tracks with sound characteristics similar to a “seed song” and features playlists according to your mood with just one click, e.g. happy/sad, calm/energetic. Recommendations are based on sound-similarity according to a variety of low, mid and high-level characteristics such as tone, color and rhythm, irrespective of metadata.

Audiogen offers a “More like this” functionality. It can seamlessly be integrated into shop solutions to provide customized music recommendations that boost music sales enormously and increase consumer loyalty.


  • Works online and offline, e.g., automatic playlist generation for streaming services, cloud services etc.
  • Musical similarity of songs in any music collection
  • Filtering of music databases by musical attributes
  • Automatic generation of musical attributes such as moods (sad, happy), instrumentation (percussions, electro-beat) etc.
  • Differentiation between speech and music


  • No lead time: recommend music without a “cold start” problem
  • Sell music on the Long Tail: find less-known alternatives to popular music and the other way around and monetize the whole back catalogue
  • >Easy connection: other data sources for rich meta-information can be linked to enhance the music discovery experience
  • Up to date: find similar songs even for new releases
  • Easy integration: Audiogen’s highly efficient architecture and low-memory footprint allows for implementation on almost any device