getting data super fast, like no waiting around

were building a new system that needs to aggregate real-time data streams from multiple sources into a single view with really low latency requirements. think sub-second. were considering using Kafka for the streams and then processing them into a database. what are peoples experiences with specific database technologies like ClickHouse, TimescaleDB, or Druid for this kind of high-throughput, low-latency aggregation? Any alternative architectures we should be looking at, perhaps?