Worked on the DataDog/datadog-agent repository to enhance the Datadog Pod Autoscaler by refining stabilization window settings across its three built-in presets. Leveraged Go and Kubernetes to reduce scale-up windows, enabling faster scaling responses while maintaining existing scale-down configurations. The technical approach involved adjusting YAML configurations, validating changes locally using minikube, Skaffold, and kubectl, and ensuring accurate profile creation with updated values. Release notes were added to document these configuration changes, and all modifications were clearly described and traceable through the associated pull request. This work focused on backend development, emphasizing reliability and responsiveness in autoscaling behavior for Kubernetes workloads.