The appointment is the clearest sign yet that the world’s biggest music company plans to build AI products in-house, not just license them.
Universal Music Group has hired one of the executives behind Spotify’s AI DJ, a strong signal that the world’s largest music company intends to build its own AI products rather than rely only on licensing deals.
Òscar Celma started Monday (Sept. 28) as senior vice president of applied AI and machine learning, reporting to chief data officer Hannah Poferl. He will oversee how AI and machine learning are developed and used across UMG’s labels and businesses worldwide.
Celma has worked on music recommendation at nearly every level of the industry. As a director of engineering at Spotify, he helped lead the teams behind AI DJ, Jam and Daylist, according to UMG. Before that, he spent seven years in senior data science roles at Pandora/SiriusXM, worked as a senior research scientist at Gracenote and co-founded Barcelona music-tech company BMAT. He joins from Condé Nast, where he ran machine learning and product analytics. He holds a PhD from Barcelona’s Universitat Pompeu Fabra and wrote the 2010 book “Music Recommendation and Discovery.”
In announcing the hire, Poferl credited Celma with building teams that turn data at scale into useful insight. Celma said he plans to advance UMG’s AI capabilities in support of human artistry. Poferl herself joined UMG in January from The New York Times, with a mandate to use data and AI to drive the company’s superfan strategy.
UMG has also invested in Stability AI, alongside Sony and Warner.
Licensing deals give a label a cut of what outside AI companies build. A senior hire with Spotify’s recommendation playbook lets UMG build tools of its own, from catalog discovery to fan-facing products and data services for its labels. For artists on Universal’s roster, the open question is how much of that technology will work for them, and on what terms.
Source: Music Business Worldwide
About the Author
