Worked on the NVIDIA/gpu-operator repository to enhance deployment flexibility and scalability for Kubernetes clusters. Developed a new CLI flag that allows configuration of the LeaderElectionNamespace, enabling the operator to support multi-namespace operation and more adaptable deployment strategies. The implementation in Go ensured that backward compatibility was maintained, so existing deployments remain unaffected when the flag is unset. This feature isolates Lease objects per namespace, supporting multi-tenant environments and improving operational stability. Focused on CLI development and Kubernetes integration, the work prioritized robust feature delivery and stability, with no major bugs addressed during this period as attention centered on this enhancement.