What do you need to think about if you want to build a centralized Ray platform for thousands of users of diverse backgrounds and use cases? Our talk shares what Spotify's ML platform team has learned and solved over the past year. We created a seamless developer experience by making it easy to not only get computational resources but also start coding on it. We enhanced our platform's reliability, scalability, performance, and cost-efficiency. We also leveraged the Ray ecosystem for ML development to solve real business problems. These efforts have led to broader adoption of Ray in ML applications at Spotify.
We describe the goals and design decisions of our managed Ray platform, our focus on a frictionless developer experience for ML practitioners, and how Ray has accelerated various ML applications. We hope our stories and learnings will inspire and help other members of the community on their Ray journey.
David Xia is a senior engineer on Spotify's ML platform team. Over the past year he helped build and operate a centralized Ray platform that enables Spotify ML practitioners to easily start prototyping their ideas and scaling their workloads. Before that he worked on Spotify's core infrastructure for backend services, specifically on deployment tooling.
Keshi Dai is a Staff ML Engineer on the Spotify Machine Learning platform team. He has been working on building and managing a centralized Ray platform to help ML practitioners at Spotify to drive business impact. He has also worked on applied ML for recommendations in the past, and has built models for ads optimization, top e-commerce websites, and music recommendation systems.
Come connect with the global community of thinkers and disruptors who are building and deploying the next generation of AI and ML applications.