Principal AI Network Engineer

About the role

Lead the networking side of large GPU clusters, from physical install and cabling through configuration, tuning and commercial readiness. Also act as top escalation for complex production network issues.

What you will do

• Lead commissioning of AI clusters: switch setup, configuration checks, host connectivity and production readiness
• Configure high-speed Ethernet or InfiniBand fabrics with RDMA/RoCE plus storage, management and out-of-band networks
• Verify topology, routing, redundancy, addressing, optics, cabling and interoperability across switches and NICs
• Build network automation, configuration management, firmware and OS baselines and repeatable deployment steps
• Run acceptance and qualification tests for link health, errors, bandwidth, latency, congestion and resilience
• Tune RDMA transport, congestion control, QoS, PFC/ECN and switch or NIC settings, and analyse NCCL results

What they look for

• Principal-level, hands-on experience bringing large-scale GPU cluster networks into production
• Deep knowledge of high-speed Ethernet and InfiniBand compute fabrics, including RDMA and RoCE
• Skill with network automation, configuration management and firmware or OS baselines
• Ability to diagnose degraded links, optical faults, packet loss and communication bottlenecks using telemetry and counters