During May 2025, Varnastadeus developed an autoscaling configuration for the cluster controller Helm chart in the castai/helm-charts repository. This work involved renaming the workloadManagement setting to workloadAutoscaling, enabling it by default, and updating permissions to manage CAST autoscaling custom resource definitions. By leveraging YAML and Helm within a Kubernetes environment, Varnastadeus enabled dynamic, demand-driven scaling for CAST AI workloads, reducing the need for manual intervention and improving resource efficiency. The implementation focused on enhancing configurability and automation in cloud infrastructure, demonstrating a solid understanding of Kubernetes resource management and Helm chart customization without introducing major bugs.