In this chapter, we learned how to process data in near real time using a streaming-based approach. Streaming processing is quite different from traditional batch-oriented processing. Through a classic word count example, we explored how streaming-oriented processing could be applied to such problems to get near real-time updates. The streaming algorithm is quite different from the classic solution to this problem, and introduces a few complex concepts, such as state management. For all the added complexity in the streaming solution, it is generally worth employing because of the significantly improved response time in gaining real-time details of the data that is being monitored.
We also looked at how to make use of the streaming-oriented approach for ML. In the next chapter, we will look at scalability concerns.
