Case Study: Musimap achieves real-time recommendations and 10x faster search with Neo4j

Neo4j Provides Musimap with a Real-Time Recommendation & Search Engine for the Music Industry

Musimap is a personalized music search and recommendation platform (founded 2014) that maps every song using 55 weighted descriptive criteria to decode the “DNA” of music. Serving clients from streaming services to promoters and restaurants, the company amassed a catalog of nearly 3 billion data items (including 50 million titles and 5 million artists) but ran into performance and scalability limits with a traditional SQL approach and needed a visual, relationship-focused model to enable real-time, nuanced recommendations.

Musimap migrated to Neo4j’s graph database, importing its catalog via a custom API to enable relationship-driven queries across rhythm, instrumentation, 400 moods and 100 contexts. The switch delivered real-time recommendation and search capabilities (searches are now over 10x faster), removed limits on dataset size, and provided high availability with automatic failover—allowing Musimap to offer complex, market-ready recommendation services 24/7.

Musimap

Pierre Lebecque

Musicology Researcher and the Creator