GPU Infrastructure Support Engineer

Role overview Provide frontline support for GPU and AI infrastructure, handling customer tickets, platform alerts, hardware diagnostics, and initial incident response across Linux systems and high-speed networks. The work includes following runbooks, coordinating on-site data center technicians, and escalating complex issues with clear evidence and handover notes.

Responsibilities - Monitor alerts and ticket queues, prioritize incidents, and respond within service targets. - Keep customers informed through clear technical updates and maintain accurate shift logs and handovers. - Diagnose GPU and server health using tools such as nvidia-smi, DCGM, dmesg, and journalctl. - Use BMC access through IPMI or Redfish for remote consoles, power actions, and event log collection. - Coordinate data center technicians for component reseats, optic replacements, and cable checks; keep asset records current during changes. - Support basic access, GPU passthrough, scheduler, and container issues, and suggest runbook improvements.

Requirements - Two to four or more years in infrastructure support, service desk, or data center operations in a customer-facing environment. - Hands-on troubleshooting experience with Linux and server hardware. - Command-line skills for systemd, permissions, storage mounts, and network diagnostics. - Working knowledge of GPU server diagnostics, including nvidia-smi and basic DCGM checks. - Familiarity with PCIe, NVMe, memory topology, and out-of-band management using IPMI or Redfish. - Understanding of incident handling, ticket prioritization, and clear written technical communication.

Nice to have - Slurm or basic Kubernetes experience. - Awareness of InfiniBand, RoCEv2, or RDMA networking. - Linux or Kubernetes foundation certifications.

Benefits and work setup - Remote work with flexibility and autonomy. - Four-day work week, except when attending events, and uncapped holiday. - Compensation and benefits are described as competitive, with no specific amounts or details provided.