Lambda Architecture uses two paths: a streaming path for fast results and a batch path for more complete results. That solves a real problem, but it creates another one. The team now has to maintain the same business logic in two places.

Kappa Architecture keeps one path. It stores source events in an append-only log, then builds views by running those events through stream-processing code. If a view is wrong, the system replays old events through the same code and rebuilds the view.

There is no separate batch layer. That sounds simple, but it only works when the event log is durable, old events can still be read, processing is safe to run again, and output stores can be rebuilt without corrupting live data.

This chapter explains what Kappa Architecture is, why teams use it, how replay replaces a batch rebuild, what the event log must guarantee, and when Kappa is a better fit than Lambda.