Apache Pinot has become a cornerstone for real-time analytics, enabling organizations to deliver low-latency insights at scale. At the forefront of this evolution recently has been Pinot’s multi-stage query engine, a transformative innovation that unlocks new possibilities for advanced analytics use cases. This session explores the journey of Pinot’s multi-stage query engine, tracing its development, key milestones, and future roadmap.

First introduced in version 1.0 primarily to enable query-time joins, the engine has since evolved to support a wide array of features such as window functions, funnel analytics, enhanced debuggability with query stats and comprehensive explain plans, and numerous performance optimizations while continuing to steadily progress toward full standard SQL semantics.

With Pinot 1.2.0 you can now utilize advanced query statistics to diagnose bottlenecks. With Pinot 1.3.0, a new multi-stage query explain plan uncovers how queries are executed. Finally, you'll discover how to leverage the reuse CTE feature recently introduced into Pinot to streamline complex queries and optimize resource utilization.

Learn practical strategies for performance optimization using the mechanics of join strategies and how to choose the most efficient approach for your use case.