Senior Infrastructure Engineer — IDC / Bare-Metal Kubernetes

About the Role

We are building our own IDC infrastructure from the ground up. As a Senior Infrastructure Engineer, you will lead the architecture, buildout, and operations of our colocation data center — from rack layout to the on-prem Kubernetes platform. You build it, you own it. This is a true greenfield role with no legacy baggage and a high degree of architectural ownership.

Your core expertise will center on three pillars: building physical infrastructure from zero, running production-grade self-hosted Kubernetes, and hybrid-cloud interconnect. As the platform matures, you'll help drive the next phase of cross-cloud and scale-out buildout.

• Design rack layout, network topology, and overall configuration for colocation

• Design and implement an isolated Out-of-Band (OOB) management network (BMC, IPMI, Redfish) with full security hardening

• Coordinate with colocation and hardware vendors — servers, switches, cross-connects, uplinks, remote hands, and cage/rack administration

• Build bare-metal automation for fleet-wide zero-touch OS provisioning (MAAS / Tinkerbell / Cluster API)

• Set up production-grade self-hosted Kubernetes clusters from scratch and maintain the core stack: Cilium (CNI), GitOps (ArgoCD), etc.

• Own cluster lifecycle: upgrades, scaling, node maintenance, incident response

• Write SOPs, runbooks, and post-mortems; participate in 24×7 on-call rotation

• Establish hybrid-cloud interconnect between IDC and public cloud (VPN / Direct Connect / peering)

• Pave the way for scale-out: GPU workloads (NVIDIA GPU Operator, passthrough, MIG) and VM workloads running alongside containers (KubeVirt or similar)

• 5+ years in data center / infrastructure / platform engineering

• Hands-on experience building physical infrastructure from scratch: rack layout, network topology, server commissioning, and coordinating cross-connects and remote hands with colo / vendors

• Practical experience designing and operating OOB management networks (BMC, IPMI, Redfish)

• Have stood up production-grade self-hosted Kubernetes from scratch, and can independently debug cluster-level issues (CNI, CSI, storage)

• Strong Linux systems administration and performance tuning (kernel, networking, storage I/O)

• Bare-metal automation experience with at least one of: MAAS, Tinkerbell, Cluster API

• Proficient with Terraform, Ansible, and at least one scripting language (Python / Go / Bash)

• Experience with Cisco network and related techniques (VLAN, LACP/LAG, BGP, ACL, etc)

• Experience with Palo Alto firewall configuration

• Experience with storage systems (NetApp, Dell EMC, Pure Storage)

• Fluent in English or Mandrain

Nice to Have

• High-density racks (30kW+) and 400G+ networking experience

• Familiarity with immutable OS (Talos Linux / Flatcar / Bottlerocket)

• Proficiency across both AWS and GCP; cross-cloud data migration experience

• Experience building storage or Bigdata / offline data clusters

• Virtualization experience (KubeVirt or similars)

• Exposure to NVIDIA GPU Operator and K8s GPU workloads

• CKA / CKS certification

• CCNP / CCIE certification

• Professional-level Japanese