This talk presents an experience-based case study of evolving a layered system into microservices and autonomous agents under cloud-native constraints.

Using a real-world example, we examine an architectural approach that improved data consistency and reliability, enabled more stable releases with minimal disruption and increased both the speed and quality of development cycles.

We explore how service boundaries were identified, which components were extracted first and when introducing autonomous agents made sense. We also cover how aggregation and reconciliation agents later improved performance, team independence and release reliability.

Finally, the session provides practical guidance on evolving layered systems incrementally, understanding when microservices are sufficient and when autonomous agents provide additional value.