State is the silent enemy of maintainable software. The more of it a system manages — scattered across tables, caches, flags, and objects — the harder it becomes to understand why the system is in a particular state. What happened? When? In what order? Event Sourcing flips this around: instead of storing the current state, you store the events that led to it. State is derived from those events — reproducible, traceable, testable.

In this talk, I'll walk you through the core concepts of Event Sourcing: events as first-class citizens of the domain, event streams, projections, and read models. Using a single running example, I'll show how you can put these concepts into practice — without specialized frameworks or dedicated event stores, but with a standard tech stack and a relational database. We'll also look at where Event Sourcing pays off, where it would be overkill — and how it gets along with Domain-Driven Design.