Use cases

ETL applications

In LorryStream, we use Streamz at the core for relaying data from streaming sources into CrateDB, because we have been looking for something smaller and more concise than Beam, Flink, or Spark. Streamz gives us the freedom to use advanced stream processing and flow control primitives within Python applications to feed databases, without the need to spin up compute clusters, or deal with Java class paths.

In this spirit, LorryStream is effectively just a humble CLI interface for Streamz.

Telemetry readings

At my company, we use it at two different points in our data processing to create a disjoint subcover of a noisy stream of telemetry readings and then blend in metadata from users and external authoritative resources to give users an understanding of what (and what kinds of) events are happening in their space.