Was good to work with HLF but the ops background required to run it in production was staggering: Replicated Kafka + ZK backing the consensus algorithm. I put it all in Kubernetes with StatefulSets. Even with my background running those systems it felt unreasonable. I quit that job in Nov so I don’t know if the situation has improved since then. I liked the architecture and the team was responsive and put care into docs. Would use again.