How should centralized exception handling be designed for production?

Central exception handling converts unexpected failures into consistent API responses while keeping endpoint code focused on normal behavior.