Supercompute Operations Engineering
Core
Own 100% GPU uptime, end-to-end, 24/7, across compute and facility for Super Intelligence stack.
Role type
Senior IC supercompute operations engineer (datacenter/facility)
Builds
GPU uptime, incident response automation, datahall deployment, and site sustainment
Domain
AI/ML infrastructure, supercomputing, datacenter operations
Deliverable
production ML models
Required skills
GPU fleet management, datacenter site operations, incident response, automation engineering, system-level datacenter knowledge, predictive maintenance, cross-functional reliability engineering, root cause analysis, hardware/firmware management, scripting (Python/Bash), systems programming
Preferred skills
Prometheus/Grafana, NVIDIA DCGM, SCADA/BMS/EPMS/DCIM, AI/ML infrastructure (GB200/NVL72/H100), InfiniBand/RoCE, fleet firmware management, playbooks/runbooks
Technologies
DCGM, Prometheus, Grafana, BMS, EPMS, DCIM, SCADA, InfiniBand, RoCE, NVLink, Redfish, IPMI, Python, Bash, C, C++, Java, Go, Rust
Responsibilities
Own monitoring architecture and automation for GPU/host/facility telemetry; manage compute hardware changes (physical/firmware/software); lead 24/7 incident response and control authority; perform day-to-day site sustainment and resolve rack/facility issues; lead acceptance testing for new data halls; develop production automation for datahall bring-up; document live datahall state; enable technician efficiency; create predictive automation for alerts/incidents; lead cross-functional reliability projects; root cause systemic failures; turn fleet data into proactive alerting and vendor corrective actions
Seniority
Senior, hands-on IC
