This tutorial track builds the third pipeline, which aggregates on the key, to produce count, average, min, and max. Since the record key is the account_id, the results are the aggregated values for each account.
The pipeline, named “payment-p3,” consumes records from the topic “converted-log1.” The predecessor from the previous tutorial track has partitioned records in topic converted-log1 based on the key. So the payment-p3 can perform group-by-key and aggregate with the correct result.
This pipeline does not call parsers and lookups. However, it has some more complicated functions. You will have to read and understand the Java code snippets.