Network Architect (AI Infrastructure)

Location: Hybrid, with presence required four days per week at the San Jose, San Francisco, or Bellevue office; Tuesday is the designated work-from-home day.

Salary: $315,000–$420,000 annually for San Francisco/San Jose; $284,000–$378,000 annually for Bellevue.

Company

Lambda builds AI cloud infrastructure for AI researchers, enterprises, and hyperscalers, with a focus on large-scale GPU computing.

What you will do

• Architect high-performance, ultra-low-latency, high-bandwidth networks for cloud platforms.
• Define network topologies and architectural patterns for GPU clusters, storage backends, and multi-tenant environments.
• Evaluate and benchmark InfiniBand, RoCE, high-speed Ethernet, Ultra Ethernet, and other next-generation networking technologies.
• Develop network architecture standards, reference designs, and scalability roadmaps for multi-site and hybrid environments.
• Partner with compute and storage architects to provide resilient end-to-end data flow and fault tolerance.
• Guide network automation, provisioning, telemetry, troubleshooting, and operational visibility while mentoring engineers and cross-functional teams.

Requirements

• 7+ years of experience architecting high-performance data center networks, preferably for HPC, AI/ML, or large-scale cloud infrastructure.
• Deep expertise in InfiniBand HDR/NDR, advanced Ethernet fabrics, RoCE, and RDMA.
• Strong knowledge of switching architectures, congestion control, QoS, VXLAN, EVPN, BGP-based fabrics, eBGP underlays, MP-BGP EVPN, ECMP, route policy, and convergence.
• Experience designing low-latency, high-throughput, redundant, fault-tolerant networks for GPU-to-GPU and storage traffic.
• Expertise in optical networking, including transceivers, fiber, link budgets, breakout architectures, and WDM.
• Experience with SDN, APIs, automation, network telemetry, and cross-functional technical leadership.

Nice to have

• AI workload profiling, NCCL, MPI, and distributed-training network tuning experience.
• Experience with network automation frameworks and telemetry tools.
• Exposure to DPU/SmartNIC technologies such as NVIDIA BlueField.
• Knowledge of multi-site interconnects, DWDM, and metro/long-haul networking.
• Experience designing customized networks with hyperscale or enterprise customers.

Culture & Benefits

• Cash and equity compensation.
• Health, dental, and vision coverage for employees and dependents.
• Wellness and commuter stipends for select roles.
• 401(k) plan with a 2% company match for USA employees.
• Flexible paid time off.