Worked on the percona/operator-lifecycle-manager repository to enhance the reliability and lifecycle management of the Catalog-Operator component. Developed a feature in Go that detects pods evicted by the kubelet due to node resource pressure, introducing a new isPodTerminatedByKubelet condition to improve pod lifecycle handling. Automated the cleanup of evicted pods to prevent stale states and resource leaks, reducing the need for manual intervention and supporting cluster stability in production Kubernetes environments. Focused on backend development and Kubernetes operator patterns, the work addressed pod lifecycle challenges under resource constraints, contributing to more robust and maintainable infrastructure management within the project.