Druid

Apache Druid is a distributed, column‑oriented data store designed for real‑time analytics. It ingests both streaming and batch data, automatically columnarizing, time‑indexing and compressing it, and provides sub‑second OLAP query latency on high‑cardinality, high‑dimensional datasets that can reach billions or trillions of rows.

The system targets developers and analysts building analytics applications that require high concurrency, low‑latency query‑on‑arrival access to events from sources such as Apache Kafka or Amazon Kinesis. It offers a SQL interface, flexible join capabilities, and configurable tiering with quality‑of‑service controls to balance performance and cost across mixed workloads.

Key architectural traits include a loosely coupled, elastic component model for ingestion, querying and orchestration, built‑in data services for backup, recovery and replication, and schema auto‑discovery that lets users ingest schemaless data while retaining type‑aware performance. The project is self‑hostable, open source under the Apache‑2.0 license, and maintained as a stable, free‑tier utility.