Jay Kreps (Co-creator of Apache Kafka® and CEO, Confluent) introduces ksqlDB, an event streaming database. As the successor to KSQL, ksqlDB seeks to unify the multiple systems involved in stream processing into a single, easy-to-use solution for building event streaming applications.
ksqlDB offers support for running connectors in an embedded mode, in addition to support for both push and pull queries. Push queries allow you to subscribe to changing query results as new events occur, while pull queries allow you to look up a particular value at a single point in time. To use a ride-sharing app as an example, there is both a continuous feed of the current position of the driver (a push query) and the ability to look up current values such as the price of the ride (a pull query).
Databases are still effective in their own realms, and ksqlDB is not intended as a replacement. Rather, ksqlDB enables you to build event streaming applications with the same ease and familiarity of building traditional applications on a relational database. It simplifies the underlying architecture for these applications so you can build powerful, real-time systems with just a few SQL statements.