Three Responsibilities of a Global Application (Part 2)

This article outlines three distinct responsibilities every global-scale application must separate: execution (doing the work), evidence/control (proving the work happened correctly), and health observability (knowing whether the system is healthy). Mixing these concerns—such as reusing CloudWatch dashboards for audit or embedding compliance logic in business code—creates systems that are hard to operate, explain, and defend. The author argues that applying a Single Responsibility principle at the architectural level, with clear separation across AWS services like Lambda, CloudTrail, and Audit Manager, makes decisions easier and systems more maintainable.