Apache Flink is a framework and distributed processing engine for stateful computations over unbounded and bounded data streams. It provides exactly-once state consistency, event-time processing and handling of late-arriving data, with layered interfaces: SQL for stream and batch data, the DataStream API and ProcessFunction for fine-grained control of time and state. Typical uses are event-driven applications, stream and batch analytics, and data pipelines or ETL. Flink is built to run in common cluster environments, with a Kubernetes Operator among its documented options, and supports high-availability setups and savepoints. It is software you deploy and operate yourself.
Flink is a project of the Apache Software Foundation, a US non-profit (501(c)(3)); its code is under the Apache License 2.0. Because there is no vendor-operated Flink service behind the project, your data is processed wherever you run the cluster: on your own servers, at an EU hosting provider or in a cloud region you select. A European organisation therefore keeps control over location and access, and takes on the operating work itself.