(Founder) Besides simple API, - Unlimited streams. Current cloud systems limit to a few thousand. With dedicated clusters, few hundred K? If you want a stream

(Founder) Besides simple API, - Unlimited streams. Current cloud systems limit to a few thousand. With dedicated clusters, few hundred K? If you want a stream per user, you are now dealing with multiple clusters. - Elastic throughput per stream (i.e. a partition in Kafka) to 125 MiBps append / 500 MiBps realtime read / unlimited in aggregate for catching up. Current systems will have you at tens. And we may grow that limit yet. We are able to live migrate streams in milliseconds while keeping pipelined writes flowing, which gives us a lot of flexibility. - Concurrency control mechanisms (https://s2.dev/docs/stream#concurrency-control https://s2.dev/docs/stream#concurrency-control)