From Silo to Collaboration - Building Tooling to Support Distributed ML Teams at Twitch
Toronto Machine Learning Series (TMLS) via YouTube
Get 20% off all career paths from fullstack to AI
Cybersecurity: Ethical Hacking Fundamentals - Self Paced Online
Overview
AI, Data Science & Cloud Certificates from Google, IBM & Meta — 40% Off
One plan covers every Professional Certificate on Coursera. 40% off your first 3 months — limited time.
Unlock All Certificates
Explore the journey of Twitch's machine learning teams as they transition from siloed operations to collaborative efforts in this insightful conference talk. Discover how Twitch addresses the challenges of distributed ML teams by developing innovative tooling and infrastructure solutions. Learn about the company's ML team structure and the strategies implemented to support efficient ML development across product areas. Gain valuable insights into Twitch's feature store, which serves as a centralized control plane while enabling distributed feature ownership. Examine the in-house ML orchestration system, Conductor, and its role in promoting best practices for pipeline management. Understand how Twitch fosters a collaborative ML culture among engineering teams, drawing parallels to community-owned open source projects. Delve into the benefits of cross-team contributions and development in enhancing ML capabilities across the platform.
Syllabus
From Silo to Collaboration - Building Tooling to Support Distributed ML Teams at Twitch
Taught by
Toronto Machine Learning Series (TMLS)