AWS EKS / AI-ML Platform Engineer

We are seeking an experienced AWS EKS / AI-ML Platform Engineer to operate, maintain, support, and continuously improve a Kubernetes-based engineering AI/ML platform running on Amazon EKS. The platform provides engineers and data scientists with an integrated environment for model development, training, workflow orchestration, distributed computing, and GPU/CPU-based workloads. The environment includes JupyterHub, Argo Workflows, Argo CD, Karpenter, Ray/KubeRay, Grafana, Prometheus, Loki, GitHub Actions, Amazon ECR, EFS, S3, RDS PostgreSQL, IAM/IRSA, KMS, and AWS networking/security services., · Manage Kubernetes namespaces, deployments, services, ingress, Helm releases, configuration, secrets, and platform services. · Coordinate EKS/Kubernetes lifecycle changes with the AWS/Cloud Operations team. · Troubleshoot pod scheduling, networking, storage, ingress, authentication, resource constraints, and application availability. · Maintain platform reliability and operational readiness.

1. JupyterHub & User Workspace Management

· Administer JupyterHub/JupyterLab environments used by engineering and AI/ML users. · Maintain user workspace configurations and container images. · Troubleshoot notebook, pod, storage, image, authentication, and resource-allocation issues. · Support CPU/GPU-based development and model-training workloads.

1. Workload Scheduling & Compute Management

· Manage Karpenter policies and Kubernetes compute provisioning. · Support dynamic CPU/GPU node provisioning and workload scheduling. · Monitor node utilization, pending workloads, scheduling failures, and capacity constraints. · Tune resource requests/limits and platform policies to balance performance, reliability, and AWS cost.

1. AI/ML & Distributed Computing Platform

· Operate Argo Workflows for ML/data workflow orchestration. · Support Ray/KubeRay distributed computing environments. · Troubleshoot distributed workloads and worker provisioning. · Support GPU workloads and associated NVIDIA Kubernetes components.

1. GitOps / CI-CD

· Administer Argo CD and GitOps-based application deployment. · Maintain integrations with GitHub / GitHub Actions and Amazon ECR. · Support controlled promotion of application/container changes across environments. · Troubleshoot failed deployments, image pulls, Helm releases, sync failures, and configuration drift.

1. Monitoring & Observability

· Maintain Prometheus, Grafana and Loki monitoring/logging capabilities. · Build operational dashboards and actionable alerts. · Monitor: o EKS cluster/platform health o Pod and node health o CPU/GPU utilization o Memory utilization o Workload queue/pending time o Karpenter provisioning o Failed workloads o Application/service availability o Storage utilization · Support root-cause analysis for platform incidents.

1. AWS Service Integration

Operate and troubleshoot platform integrations involving: · Amazon EKS · Amazon ECR · Amazon EFS · Amazon S3 · Amazon RDS for PostgreSQL · Application Load Balancer / AWS Load Balancer Controller · Route 53 · AWS WAF · AWS Certificate Manager · IAM / IRSA · AWS KMS · AWS Systems Manager Parameter Store · VPC endpoints / private connectivity

1. Platform Lifecycle & Upgrades

This should be a major responsibility of the role, rather than an incidental support activity. · Own lifecycle management of platform Helm charts and application components. · Plan, test, and execute platform upgrades. · Track compatibility between EKS/Kubernetes and Karpenter, Argo, JupyterHub, KubeRay, Ray, PrometheGrafana and supporting components. · Maintain upgrade runbooks and rollback procedures. · Coordinate platform upgrades with AWS infrastructure teams and application owners.

1. Security & Access

· Support OIDC-based application authentication. · Manage Kubernetes RBAC and IAM Roles for Service Accounts (IRSA). · Maintain secrets and configuration securely. · Support encryption and certificate management. · Participate in vulnerability remediation and platform patching.

1. Production Support

· Provide operational support for platform incidents. · Perform root-cause analysis and corrective actions. · Develop operational runbooks and troubleshooting procedures. · Automate repetitive administrative tasks. · Work with Cloud Operations, application teams, engineering users, AWS, and other technology teams to resolve complex problems., · Strong hands-on Kubernetes / Amazon EKS administration · AWS infrastructure knowledge · Helm · Argo CD · Argo Workflows · Karpenter · JupyterHub / JupyterLab · Kubernetes ingress and AWS Load Balancer Controller · Prometheus / Grafana · Git / GitHub and CI/CD · Docker/container technologies · Linux administration · Kubernetes networking and storage · IAM / IRSA · ECR, EFS, S3 · Infrastructure troubleshooting · Python and/or Bash automation

Recommended level: Senior Platform Engineer / Lead Platform Engineer I would target approximately: · 7+ years infrastructure/platform/DevOps experience · 3+ years hands-on Kubernetes experience · 2+ years supporting production Amazon EKS environments · Demonstrated experience operating production Kubernetes platforms rather than only deploying applications onto Kubernetes