I just wrote this blog post on a very interesting pattern:
→ Stream with raw data (billions) → Filter → Stream of insights (Millions) → Connector → Workflow to take actions (Thousands)
seen in real-life: Data Streams Are Nothing Without Actionable Insights Leading to Real-Life Actions
Happy to discuss!
Bernd
Great post @berndruecker. I’m curious if you think that Camunda and Kafka are a good combination for solving strict workflow solutions. By strict I mean typical “when this do that” scenarios. In the past I’ve struggled with how to handle processing failures in workflow steps and Kafka’s messaging durability presented some interesting design challenges when re-invoking failed tasks or handling manual intervention. Maybe Camunda has some interesting features to assist with these conerns?
Thanks for contributing to the Confluent Blog!