streams intro

04-kafka-streams/README.md

@@ -1,11 +1,47 @@
 # Kafka Streams API
 
+Kafka Streams is a library that can be used to enable stream processing 
+capabilities on your applications.
+
+This means that you can process data "on the fly" (i.e. subscribe to 
+events on Kafka) and producer new events. This sounds very similar to
+Producer and Consumer API, and it is because those are used inside 
+Kafka Streams. But plus consuming and producing events, the capabilities to process
+the stream is what makes Streams a powerful API.
+
 ## Streams Topology
 
-## DSL: Stateless Operations
+There are 2 APIs in Kafka Streams: 
+
+* Processor API
+* Streams DSL
+
+Where the API define `Processor` stages where data that comes from 
+Sources (i.e. Topics) are processed.
+
+But we have a DSL that simplifies the processing of data, and have common 
+operators to process data (e.g. mapping, filtering, joining, etc.)
+
+The goal of these APIs is to define a `Topology` that represents 
+the Stream Application.
+
+![](stream-process.png)
+
+## Streams DSL
+
+Inside Kafka Streams DSL API, there are 2 main abstractions:
+
+* `KStreams`
+* `KTable`
+
+`KStreams` represents data in motion and `KTable` represents data at rest.
+
+Both notions are key to implement Streaming Applications.
+
+### DSL: Stateless Operations
 
-## DSL: Stateful Operations
+### DSL: Stateful Operations
 
-## Tables and Joins
+### Tables and Joins
 
 ## Windows