Secure Kubernetes Platform Engineering for EKS and GKE Helm Based Standardization and Runtime Governance

Many organizations run their applications on managed Kubernetes platforms such as Amazon EKS and Google GKE. These services reduce infrastructure management effort, but security problems still occur at the workload level. Configuration drift, unsafe defaults, and inconsistent controls across environments often introduce vulnerabilities. This study presents a structured platform engineering model that combines Helm-based deployment standardization with Policy-as-Code runtime governance in a multi cloud setup. Controlled experiments were conducted on parallel EKS and GKE clusters. The evaluation measured deployment consistency, drift frequency, misconfiguration reduction, and policy enforcement performance. After Helm standardization, deployment consistency increased from 78.3% to 99.2%. Weekly configuration drift events decreased by approximately 90%. Runtime policy controls blocked 98.6% of simulated violations. Average admission latency remained below 25 milliseconds under normal workload conditions and below 40 milliseconds under peak load. Security metric differences between EKS and GKE remained under 1%. The findings indicate that consistent deployment templates combined with admission based policy validation significantly reduce configuration risk while maintaining acceptable system performance in multi cloud Kubernetes environments