In modern cloud architectures, building resilient systems goes beyond application-level retries and logging. Resource-level failures — like an EMR cluster terminating unexpectedly — can silently disrupt pipelines if left unmonitored. Leveraging AWS CloudTrail, EventBridge, and lifecycle events, our approach captures critical infrastructure signals, tags resources with meaningful attributes, and triggers automated workflows for alerts, metadata updates, and recovery. This event-driven design transforms transient failures into actionable insights, enabling robust, self-healing systems and bridging the gaps traditional monitoring often misses.