Summary
Joan Serrà is a seasoned research scientist and team lead specializing in machine learning for audio and multimodal data, currently heading SonyAI's EU research team. With over 20 years spanning academia and industry and a PhD from Universitat Pompeu Fabra, he blends deep theoretical expertise with product-focused research in generative modeling, representation learning, and audio retrieval. He has led multi-geo teams at Dolby and Telefónica, published extensively, and regularly teaches advanced deep learning topics at universities. Known for a strategic and creative technology vision, Joan combines hands-on model development with mentorship and cross-disciplinary collaboration, often moving concepts from papers to deployable systems. An often overlooked strength is his long-running focus on time-series and music analysis, which gives him uncommon depth in temporal generative models for audio.
9 years of coding experience
19 years of employment as a software developer
Undergraduate Electrical and Electronics Engineering, Undergraduate Electrical and Electronics Engineering at Universitat Ramon Llull
PhD Information and Communication Technologies, PhD Information and Communication Technologies at Universitat Pompeu Fabra
English, Spanish, Catalan