Persona Music, a U.S.-based production music library, partnered with AudioShake to unlock deeper value from its catalog of human-created music. Like many music libraries, Persona had a vast archive of mixed audio—high-quality, finished tracks that were difficult to repurpose at scale due to their flattened format.

Using AudioShake’s state-of-the-art stem separation technology, Persona was able to transform these stereo mixes into structured, multi-stream datasets, isolating vocals, drums, bass, and other instruments into individual stems. These clean components became the foundation for new creative workflows—including the development of generative music models trained on Persona’s own audio DNA.

By converting its existing content into training-ready data, Persona enabled its team to generate new permutations and adaptive versions of its tracks while maintaining the integrity of the original recordings. Crucially, this approach also preserved attribution and royalty pathways, ensuring original artists remained properly credited and compensated.

‍