System Development Engineer – DRAM Memory Optimization

About the role

As a member of the Region Flexibility Migration (RFM) team, you will lead Amazon's DRAM memory optimization initiative for EKS clusters. You will own the full lifecycle of cluster provisioning, memory profiling, and optimization to reduce the DRAM footprint across Amazon's Kubernetes infrastructure.

Key responsibilities

• Provision, configure, and maintain EKS clusters, including node groups, autoscaler tuning, and Fargate profiles with memory‑optimized settings.
• Perform deep memory profiling of containerized workloads, identify over‑provisioned requests/limits, detect leaks, and analyze kernel and cgroup memory usage.
• Design, implement, and validate memory‑optimization strategies such as VPA right‑sizing, node‑pool instance type selection, and memory‑efficient scheduling policies.
• Run custom experiments using Amazon internal tooling, including load, soak, and chaos engineering tests, to measure impact and ensure reliability.
• Collaborate with service teams to review cluster architectures, propose optimization plans, and communicate findings to engineering leadership.
• Create runbooks, SOPs, and technical documentation that capture optimization patterns for both engineers and AI‑driven automation.

Required profile

• Self‑driven professional comfortable managing end‑to‑end EKS operations.
• Strong analytical mindset with experience in memory analysis of distributed micro‑services.
• Ability to design experiments, interpret metrics, and derive actionable insights.
• Effective communicator who can work across multiple engineering teams and leadership.

Required skills

• Amazon Elastic Kubernetes Service (EKS)
• Kubernetes cluster management
• Container runtime (Docker) memory profiling
• Vertical Pod Autoscaler (VPA) configuration
• AWS services and instance type optimization
• Fargate profiles
• Load testing, soak testing, and chaos engineering
• Linux kernel memory accounting and cgroup limits