about
I'm working at Spotify as Director of Engineering for the Listening Experiences.
Before that, Pandora / SiriusXM as VP of Data Science, where I managed a team of scientists and musicologists.
Also a Senior Research Engineer at Gracenote, and Chief Innovation Officer at BMAT
You can reach me on:
book
some press
- 2020: How to Make the Perfect Playlist by New York Times [en]
- 2019: How Pandora knows what you want to hear next? by Forbes [en]
- 2017: Los 20 latinos más influyentes de la indústria tecnológica en 2017 by CNET [es]
- 2017: Pandora knows you're secretly not that hip by VentureBeat [en]
- 2017: Pandora uses Machine Learning to make sense of 80 billion thumb votes by VentureBeat [en]
- 2017: How Pandora built a better recommendation engine by The Server Side [en]
publications
Music Recommendation Datasets (from Last.fm)
a selection of invited talks
- Exec(ut) (Utrecht, NE) [en]: 5 Lessons Learned Building a Music Recommender System (2019) [video]
- World Summit AI (Amsterdam, NE) [en]: Lessons Learned Building a Large Recommender System (2018) [video]
- SXSW (Austin, TX) [en]: Machine Learning: The Art of Explore vs. Exploit (2017)
- Mobile Beat (San Francisco, US) [en]: The Feedback loop: AI, Machine Learning and the Human in the Middle (2017)
- Developer Week (San Francisco, US) [en]: Music Artificial Intelligence with a Human Touch (2017)
- Data Science Summit (San Francisco, US) [en]: The Explore-Exploit Dilemma in Music Recommendation (2016)
- Recommender Systems Conference (Boston, US) [en]: Inside Pandora: The Explore-Exploit Dilemma in Music Recommendation (2016)
- ML Conf (New York, US) [en]: Music Recommendation and Discovery at Pandora (2015)
- MIA Music Summit (Miami, US) [en]: Music Discovery: Mix of Art and Science (2015)
- Strata+Hadoop (San Jose, US) [en]: Inside Pandora: Practical Application of Big Data in Music (2015)
- QConSF (San Francisco, US) [en]: Inside Pandora: Ten Years After (2014)
- The Hive Think Tank (Oakland, US) [en]: The Data Driven Juke Box (2014)
- Recommender Stammtisch, Soundcloud (Berlin, Germany) [en]: MusicDiscoBerry: the bittersweet high-hanging fruit (2012)
- Stanford University, CCRMA (Palo Alto, US) [en]: Music Similarity is Dead. Long live Music Recommendation! (2012)
- Information Theory and Applications (San Diego, US) [en]: Music Similarity is Dead. Long live Music Recommendation! (2012)
- MIDEM (Cannes, France) [en]: The Future of Music Consumption (2011)
- Future Music Forum (Barcelona, Spain) [en]: Digital Music Innovation (2010)
- SFMusicTech (San Francisco, US) [en]: Machine Listening Panel [audio] (2010)
toy projects
- python-RecSys : A python library for implementing a recommender system
- python-iTunes : Python wrapper for iTunes Store API (over 2,000 downloads at pypi)
- PheXist : A PHP5 class to XQuery eXist native XML:DB (over 4,000 downloads)
- Eerfy (New York, 2013): Solving the Music Industry Problem in 24h. more info
- HellaBar (Oakland, 2013): recommends bars, restaurants, cafes, pubs & clubs based on the music you like more info
- The Silent Treatment ("Hollywood Hack Day", Los Angeles, 2013): convert any movie to a silent film more info
- Face The Music (Barcelona, 2013): Fig. to accept the (un)pleasant results of one's
actionsface more info - [EXPLICIT] feedback (Cannes, 2013): Love/ban a track in a playlist using hand gestures with your webcam more info
- LED there be rock (Stockholm, 2012): Control XMAS tree lights with Arduino + audio more info
- Buddhafy (San Francisco, 2012): Control a Spotify playlist with your mind! slides
- Groovify (San Francisco, 2010): a twitter bot that recommends you music that you're #nowplaying
- GeoTracklist (2010): a playlist generator based on current geolocation
- Route 666 (Barcelona, 2010): a road-trip playlist generator
- Kuku! (Barcelona, 2010): Kukuxumusu visualization of a last.fm user profile
- Genre navigator (2009): Tag Cloud of the main Last.fm genres
biography
- 2021-: Director of Engineering at Spotify [en]
- 2019-2021: VP of Data Science at Pandora / SiriusXM [en]
- 2017-2019: Head of Research at Pandora [en]
- 2017: Top-20 Latinos in Tech by CNET [es]
- 2014-2017: Director of Research at Pandora [en]
- 2011-2014: Senior Research Engineer at Gracenote [en]
- 2008-2011: Chief Innovation Officer at BMAT [en]
- 2010-2011: Torres y Quevedo Post-doc R&D Research Grant at BMAT [en]
- 2010: Springer Book "Music Recommendation and Discovery: The Long Tail, Long Fail and Long Play in the Digital Music Space" [en]
- 2009: Co-founder of EuroMM, the European Chapter of the SIGMM [en]
- 2008: PhD, "Music Recommendation and Discovery in the Long Tail", Department of Information and Communication Technologies, UPF [en]
- 2007: Research grant from the Mexican government. Invited researcher in the Computer Science Research Center (CIC-IPN) [es]
- 2006: 2nd prize of the Semantic Web Challenge [en], for the Foafing the Music system
- 2006: Invited expert of the W3C Incubator Group - Multimedia Semantics [en]
- 2005: Co-founder of the BMAT company [en], a spin-off of the MTG
- 2004: Professor Col.laborador (Assistant Professor), from the Agencia de Qualitat Universitaria (AQUCatalunya) [ca]
- 2003-2009: Teaching at the Universitat Oberta de Catalunya [ca]
- 2001-2008: Teaching at the Universitat Pompeu Fabra (UPF) [ca]
- 2000-2008: Research Scientist and Project Manager at the Music Technology Group (MTG) [en]
tutorials
- ACM RecSys 2011 (Chicago, US) [en], together with Paul Lamere: "Music Recommendation and Discovery Revisited" (2011) [slides]
- ACM Multimedia 2008 (Vancouver, Canada) [en]: "If you like The Beatles you might like...A Tutorial on Music Recommendation" (2008)
- ISMIR 2007 (Vienna, Austria) [en]: "Music Recommendation Tutorial" [en], together with Paul Lamere (2007) [slides, pdf]
- MICAI 2007 (Aguascalientes, Mexico) [en]: "Tutorial on Recommender Systems" (2007)