Worked on enhancing the scale-up reliability and maintainability of the rancher/autoscaler project by implementing asynchronous node group creation within the Scale-Up Orchestrator. This approach enabled parallel provisioning, reducing scale-up time and improving the robustness of the scaling process. Refactored core orchestrator logic for greater reusability and maintainability, while consolidating error handling to provide clearer user-facing feedback during concurrent scale-ups. Removed redundant test code to streamline ongoing maintenance and support future scalability. The work leveraged Go for backend development and utilized cloud computing concepts alongside Kubernetes, focusing on building a more reliable and scalable infrastructure for automated cluster provisioning.