Diagnosing Performance Bottlenecks in Databricks Lakehouse

In complex analytics environments, performance issues are rarely isolated — and surface-level tuning often misses the real bottleneck. This deep dive session presents an advanced framework for diagnosing performance problems in Databricks Lakehouse, grounded in how the engine executes queries at scale. We’ll walk through how to distinguish query-level inefficiencies from data layout issues, concurrency pressure, and system-level constraints, and how to interpret execution plans and runtime signals to identify the true root cause. Attendees will leave with a repeatable diagnostic approach they can apply to large, multi-tenant production workloads and examples of using AI to assist in finding and resolving your performance bottlenecks.