Serverless Compute for Notebooks, Jobs, and Pipelines: Latest Launches and What's Next

Serverless Compute for Notebooks, Jobs, and Pipelines has reached a turning point. What started as a convenience for ad-hoc and smaller workloads now runs petabyte-scale production and mission-critical SLAs at some of the largest Databricks customers. With Standard mode for Jobs and Pipelines, serverless is cost-competitive with well-tuned classic compute while removing the operational burden of cluster sizing, autoscaling, and runtime upgrades.This session shares concrete lessons from customers running serverless at serious scale, covers the most recent launches across runtime, performance, and cost controls, and previews what is coming next in workload compatibility and governance.If you have been waiting for serverless to be ready for your hardest and largest workloads, this is the session to recalibrate.