Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore how Apache Pinot enables native upsert support in real-time OLAP systems without compromising performance in this 29-minute technical conference talk. Discover the architectural innovations that allow Pinot to handle billions of primary keys per node while maintaining ingestion freshness SLAs of seconds and processing hundreds of thousands of events per second. Learn how traditional OLAP systems struggle with mutable data and real-time requirements, and examine Pinot's solutions for modern analytics use cases including ride-sharing demand prediction and real-time customer profile enrichment. Master techniques for efficient primary key management at scale, fast recovery mechanisms, strict operational SLA maintenance, and cost optimization strategies as tables grow dramatically. Understand the core design principles behind Pinot's native upsert implementation and gain practical insights for building real-time analytics systems on constantly evolving datasets that deliver both data freshness and query flexibility while maintaining optimal speed and cost efficiency.
