Senior Network Engineer - InfiniBand / UFM

Who We Are

Lightning AI is the company behind PyTorch Lightning. Founded in 2019, we build an end-to-end platform for developing, training, and deploying AI systems—designed to take ideas from research to production with less friction.

Through our merger with Voltage Park, a neocloud and AI Factory, Lightning AI combines developer-first software with cost-efficient, large-scale compute. Teams get the tools they need for experimentation, training, and production inference, with security, observability, and control built in.

We serve solo researchers, startups, and large enterprises. Lightning AI operates globally with offices in New York City, San Francisco, Seattle, and London, and is backed by Coatue, Index Ventures, Bain Capital Ventures, and Firstminute.

The Role

We are seeking an experienced Senior Network Engineer (InfiniBand / UFM) to design, deploy, automate, and operate next-generation AI Factory networking infrastructure supporting large-scale GPU clusters. This role is responsible for building and maintaining high-performance NVIDIA Quantum InfiniBand fabrics that power AI training and inference environments utilizing NVIDIA UFM, NCCL, RoCEv2, and modern data center technologies.

The ideal candidate has deep expertise in InfiniBand networking, NVIDIA Unified Fabric Manager (UFM), large-scale GPU deployments, Linux networking, automation, and troubleshooting distributed AI workloads.

What You’ll Do

• Design, deploy, and maintain large-scale NVIDIA InfiniBand fabrics supporting AI/ML GPU clusters.
• Deploy and administer NVIDIA Unified Fabric Manager (UFM) Enterprise for monitoring, provisioning, telemetry, and fabric health.
• Configure and optimize NVIDIA Quantum and Quantum-2 InfiniBand switches.
• Troubleshoot fabric performance issues impacting NCCL, MPI, GPUDirect RDMA, and AI training jobs.
• Implement and validate fat-tree, Dragonfly+, Clos, and spine-leaf network architectures.
• Perform firmware lifecycle management for InfiniBand switches, adapters (HCAs), and UFM infrastructure.
• Optimize congestion control, adaptive routing, QoS, and traffic engineering for high-performance GPU communication.
• Work closely with AI platform, GPU infrastructure, storage, and systems engineering teams to deploy scalable AI Factory environments.
• Automate network provisioning using Python, Ansible, Git, REST APIs, and Infrastructure-as-Code methodologies.
• Monitor network health using UFM telemetry, Prometheus, Grafana, and other observability platforms.
• Support high availability, maintenance windows, incident response, root cause analysis, and capacity planning.
• Participate in architecture reviews and define networking standards for AI infrastructure.

Required Qualifications

• 7+ years of data center networking experience.
• 3+ years supporting NVIDIA InfiniBand environments.
• Hands-on experience with NVIDIA UFM Enterprise.
• Experience deploying and operating Quantum and Quantum-2 InfiniBand switches.
• Strong understanding of:
  • InfiniBand Architecture
  • Subnet Manager (SM)
  • Adaptive Routing
  • Congestion Control
  • Partition Keys (PKeys)
  • LIDs
  • Queue Pairs (QP)
  • Virtual Lanes (VL)
  • Service Levels (SL)
• Strong Linux administration experience (Ubuntu).
• Experience with automation using Python and Ansible.
• Deep understanding of Layer 2 and Layer 3 networking.
• Experience with BGP, EVPN, VXLAN, and modern spine-leaf architectures.
• Experience with packet captures and troubleshooting using tcpdump, Wireshark, and ibdiagnet tools.
• Excellent troubleshooting and communication skills.

What You Bring

• 10+ years of experience in large-scale data center networking
• Deep expertise in spine-leaf architectures and L3 fabrics
• Strong experience with BGP, EVPN, VXLAN
• Experience operating high-performance computing (HPC) or GPU-dense environments
• Experience designing networks for hyperscalers, neoclouds, or high-scale SaaS infrastructure
• Strong automation background (Python, Ansible, Terraform, or similar)
• Experience with network observability tooling and telemetry pipelines
  Proven ability to design systems that scale to thousands of nodes
• Strong documentation and architectural communication skills

Nice to Have

• Experience with Netris and Terraform.
• Experience with multi-region backbone design
• Exposure to bare-metal provisioning systems
• Experience working in high-growth infrastructure startups

**Why This…