Inside Spotify’s Machine Learning Leadership
Tony Jebara joined Spotify New York last summer as VP of Engineering and Head of Machine Learning. He splits his time between two distinct responsibilities: running the engineering organization behind personalization and steering the company-wide machine learning strategy. Here’s how he describes the rhythm of his work.
A Dual Role in Engineering and ML
Jebara’s “vertical” role puts him in charge of the engineering teams that build personalization across the product. That means ownership of the Home page, Search functionality, Voice interaction, and the Programming Platform handling content and collections. Roughly 115 people report to him, mostly based in New York, Boston, and Stockholm. These teams operate with clearly defined roadmaps and quarterly delivery targets, so much of his time goes to coordinating with Product Managers and other stakeholders to keep everything on schedule.
His “horizontal” role as Head of Machine Learning spans the broader company. He leads several key work groups, notably the Machine Learning Tech Strategy group, which brings together representatives from many business units to decide where Spotify invests in ML and which cross-functional capabilities to strengthen.
One current focus is improving the correlation between online and offline behavior—ensuring that prototypes developed in offline settings perform consistently when deployed into the live environment. That effort is part of a company-wide initiative to advance machine learning at Spotify.
Typical Days and Constant Collaboration
His schedule is dense. Mornings begin with getting his three young children ready for school, followed by a commute downtown during which he checks email and Slack. The bulk of the day is filled with meetings, hangouts, and one-to-ones across both of his roles. Lunch is often a quick trip to the office salad bar, with occasional time carved out to discuss hiring strategy with recruitment or HR teams.
Beyond internal work, Jebara makes time for external relationship building. He meets regularly with partners at companies like Google and with the SoundBoard council companies that advertise on Spotify. He also represents Spotify’s machine learning work at conferences and, every few weeks, invites outside speakers to deliver seminars at the office to keep the team current on industry developments.
“I’ve been really impressed by how open-minded and curious people are here. They’re receptive to technical conversations and want to understand new ways of working.”
Culture and Work-Life Balance
Despite being relatively new to the company, Jebara emphasizes the collaborative culture he has found. People are willing to take on new challenges and share knowledge, which fosters innovation and a stronger connection with listeners. He describes it as a place where “everyone’s working together to help everyone else succeed”—a feeling akin to being part of a band.
His afternoons usually wind down around 5:00 or 5:30 p.m., letting him get home for his children’s dinner. Evenings often include an hour or two of emails and reading once the kids are in bed. Weekends are reserved for road trips out of the city with his family, a deliberate change of pace from a week that rarely leaves gaps in his calendar.
