Primary difference is Confluent includes capabilities specific to Kafka for running it at scale. Replicator to replicating data across data centers. auto data balancer for managing/tuning partitions, and Control Center for streams monitoring and monitoring/managing cluster health. In addition, Confluent supports the full stack for Kafka, specifically Connect and Streams, which Hortonworks does not. So if you're running either of these two and something goes wrong, Hortonworks will not help you, you're on your own. Finally, Confluent has unmatched expertise in Kafka. As the team that created the project, they have a unique understanding of what it takes to run it at scale and the engineering breadth across the core project.