EKS Cluster Operations Scaling Beyond Team's Manual Capacity
Your team at a regional insurance company runs production workloads for multiple business units on Amazon EKS clusters. The platform hosts digital applications and core insurance solutions serving customers across all touchpoints. Over the past 18 months, the number of containerized microservices has grown significantly, with teams from several distinct business units onboarding their applications. Your DevOps and Cloud team — which remains intentionally lean — is now spending a disproportionate amount of time on recurring infrastructure tasks: manual node patching, coordinating quarterly Kubernetes version upgrades, and tuning cluster scaling configurations. Incidents are occurring during routine maintenance windows because node replacements sometimes terminate pods from the same microservice simultaneously. Application teams report delays in getting infrastructure support, and the ops team flags that they are unable to keep pace with both maintenance obligations and new onboarding requests at current staffing levels. Before your team commits to the next round of hiring to handle infrastructure overhead, what architectural approach would you recommend to the leadership review?