Looking for patterns with Flink CEP library
Flink provides a Complex Event Processing (CEP) library for quite some time already. It satisfies needs of many applications in areas like click stream e.g. reacting to user interactions or financial sector e.g. handling stock values changes. Nevertheless recently it was under heavy development, introducing lots of improvements and new features. In this talk I will discuss how newly introduced pattern categories like optionals, discarding and counting patterns or kleene closures expands the spectrum of possible use cases even further. With the examples inspired by music streaming platforms I will show how you can use Flink CEP library to specify and detect dynamic patterns in your real-time streaming applications. I will also describe mechanisms of underlying NFA(Non-finite automaton) that allowed those changes.