Stream processing was born out of a desire for real-time analytics. But modern stream processors, like Apache Flink, have progressed to the point where it is now practical and interesting to use them for building event-driven applications and services.
This talk will provide an introduction to Apache Flink, showing what it's like to build applications on a platform that's based on data streams, state, and snapshots. Flink manages an application's code and associated state together as a set of tightly integrated, scalable units, in contrast to the layering that separates computation from storage in traditional systems. This approach to building event-driven applications has many desirable properties, including its simplicity, robustness, and its ability to provide high throughput event processing with low latency.
A Talk by David Anderson, User Education at Data Artisans.