Netflix on AWS

Netflix operates actively across four AWS Regions, serving global traffic by intelligently directing users and managing costs through thousands of auto-scaling compute server groups. At times, traffic surges hit their service that could exceed capacity. This session walks through how Netflix solves these problems by combining predictive automated pre-scaling with fast reactive auto-scaling, along with advanced resilience techniques. These techniques include prioritized load shedding, cross-Region traffic shifting, targeted capacity injection based on service criticality, and more. Learn how to maintain low operational costs while being prepared for sudden load spikes.

Data science at Netflix goes far beyond eponymous recommendation systems and touches every aspect of the business, from optimizing content delivery to fighting fraud. Netflix’s unique culture affords its data scientists extraordinary freedom of choice in tools, which results in an ever-expanding set of machine learning approaches and systems. In 2019, Netflix open-sourced its human-centric ML platform, Metaflow. In this session, Netflix shares some lessons learned in its multi-year journey building the ML systems that Metaflow incorporates.