[mohinish]

Companies have built enormous batch processing systems to transform, format, and store data in systems that are ready to be queried by applications across their organizations. Batch processing is thus a result of the limitations of SQL since the data has to be queried repeatedly to keep it fresh.

What if the same data is processed in real-time as it flows in with applications instantly able to access, minus all the complexity of maintaining a batch processing pipeline? I believe this is going to be the promise of streaming database systems.