Worked on enhancing Kubernetes scheduling performance for large-scale deployments by contributing to both the kubernetes/kubernetes and kubernetes/dynamic-resource-allocation repositories. Focused on reducing lock contention and restructuring resource access, the developer implemented optimizations such as splitting resource slices into Shared and OnNode categories to accelerate scheduling decisions and minimize filter-stage latency. Leveraging Go, concurrent programming, and deep knowledge of Kubernetes scheduler plugin architecture, these changes improved throughput and responsiveness for clusters with thousands of pods and nodes. The work established scalable cross-repo optimization patterns, resulting in more efficient CPU utilization and reduced wait times during peak scheduling workloads in backend systems.