Data Center Engineer — Kubernetes

Role overview

Design, build, and operate the infrastructure that connects physical data-center machines, AWS EKS, Kubernetes clusters, and PERCO’s worker system into a reliable deployment platform.

Responsibilities

• Design data-center deployment solutions across compute, storage, networking, power, access, and operational constraints
• Provision, commission, inventory, and troubleshoot physical Linux servers and their network paths
• Build and operate Kubernetes clusters across AWS EKS and self-managed environments
• Integrate clusters and machines with PERCO’s worker system for controlled scheduling, execution, status reporting, and lifecycle management
• Automate provisioning, upgrades, scaling, recovery, and compliance checks using Infrastructure as Code
• Establish observability for cluster health, workloads, capacity, networking, and deployment evidence
• Diagnose production failures across hardware, Linux, networking, Kubernetes, cloud services, and application workloads
• Write runbooks and ensure incidents end with root-cause fixes or clearly owned follow-ups

Requirements (signal)

• Proven production experience operating Kubernetes and Linux infrastructure
• Hands-on experience with AWS and EKS (networking, identity, storage, cluster lifecycle)
• Ability to provision and troubleshoot physical or virtual machines, networking, DNS, certificates, and container runtimes
• Experience automating infrastructure with Terraform, Ansible, Helm, or equivalent tools
• Strong incident-debugging skills using logs, metrics, events, and system state
• Ability to communicate architecture decisions, operating risks, and recovery evidence clearly
• Willingness and ability to travel to U.S. data centers or customer sites as needed
• Authorization to work in the U.S. when employment begins (OPT allowed; H-1B sponsorship supported for eligible candidates)

Bonus (nice to have)

• Bare-metal Kubernetes, Cluster API, Talos Linux, PXE/image-based provisioning, or immutable host systems
• GPU infrastructure, workload scheduling, or distributed worker systems
• Hybrid-cloud networking (VPN/private connectivity) and multi-cluster operations
• Designing secure customer-managed or regulated deployment environments

Location / work model

• Remote-first (based in the United States)
• Option to work from the San Francisco, CA area
• Planned travel required for physical deployment, commissioning, or incident response.