Apache Druid is a real-time analytics database for interactive data applications. Its distributed architecture can ingest millions of events per second and execute sub-second queries at virtually limitless scale for rapid data exploration. It supports both streaming and batch data ingest - with the unique ability to instantly query real-time data from Kafka and Kinesis and correlate against historical data. With Kafka as the foundation for streaming data, Druid complements the data architecture with high performance, interactive analytics at scale to deliver real-time, situational intelligence. This talk shows real-world experiences building out real-time analytics stacks powering next-generation observability, customer insights, and insights as a service.