how to process data in chunks/batches with kafka streams? Doesn't seem to exist yet. Watch this space https://issues.apache.org/jira/browse/KAFKA-7432