In the implementation yes, minimizing single-event latency and maximizing batch throughput are largely mutually incompatible. But from a user/API standpoint, th

In the implementation yes, minimizing single-event latency and maximizing batch throughput are largely mutually incompatible. But from a user/API standpoint, there's little difference.