Shiva developed the Driver Pod Resource Configuration feature for the NVIDIA/gpu-operator repository, enabling configurable resource requests and limits for all driver pod containers, including driver, GDS, GDRCopy, and the driver toolkit. Using Go and YAML within a Kubernetes environment, Shiva implemented a consistent approach to resource governance by leveraging config.Driver.Resources, allowing explicit quality of service and resource guarantees. This work improved predictability, stability, and scalability for GPU workloads in production clusters by supporting better capacity planning and performance isolation. The feature demonstrated a focused, in-depth engineering effort, addressing a core operational challenge in cloud native GPU management.