How do you handle database sharding and replication in a system that experiences high write volumes at Wise. Energy?
