Explore the challenges and solutions in building effective MLOps pipelines for autonomous factories in this 34-minute conference talk by Saranyan Vigraham, Head of Engineering at Petuum. Discover how Petuum developed flexible ML pipelines to deploy hundreds of models in dynamic factory environments. Learn about design patterns, architectural approaches, and anti-patterns based on real-world experiences. Gain insights into the unique challenges of MLOps compared to DevOps, and understand how to create pipelines that allow for easy iteration and concurrent model testing. Benefit from Saranyan's expertise in leading engineering teams and his experience in shipping enterprise and consumer products at companies like Meta and PayPal.
MLOps That Works: Building ML Pipelines for Autonomous Factory Models
Toronto Machine Learning Series (TMLS) via YouTube
Overview
Syllabus
Saranyan - MLOps that works: How we built ML pipelines for deploying models for autonomous factories
Taught by
Toronto Machine Learning Series (TMLS)
