Source Job

  • Lead network fabric bring-up and integration for large-scale AI clusters, including switch initialization, RDMA/RoCE configuration, and host connectivity validation.
  • Execute performance optimization and acceptance testing for link health, bandwidth, latency, congestion control, and multi-node NCCL distributed workload performance.
  • Serve as Tier 3 escalation for production incidents, driving root cause analysis, network reliability improvements, and ongoing performance tuning.

InfiniBand Network Automation

12 jobs similar to Principal AI Network Engineer

Jobs ranked by similarity.

US

  • Design, deploy, and maintain InfiniBand and RoCE network fabrics for GPU training and inference.
  • Manage data center switching, routing, and customer connectivity solutions.
  • Drive automation, capacity planning, and incident resolution across high-performance networks.

The hiring company builds and operates high-performance networks for GPU computing and AI workloads. Its infrastructure team focuses on scalability, reliability, and cutting-edge networking technologies.

Europe

  • Design, deploy, and maintain high-performance network infrastructures for HPC environments with a focus on InfiniBand fabrics.
  • Troubleshoot complex network issues across InfiniBand and Ethernet environments.
  • Manage and optimize InfiniBand components, including switches, HCAs, subnet managers, and fabric configurations.

Mirantis is the Kubernetes-native AI infrastructure company, enabling organizations to build and operate scalable, secure, and sovereign infrastructure for modern AI and data-intensive applications. It serves many of the world’s leading enterprises and is committed to open standards and freedom from lock-in.

US Unlimited PTO

  • Design and bring up networks in large-scale GPU clusters, including compute fabric, storage, front-end, and out-of-band management.
  • Own cabling plans, optics BOMs, and labeling standards; lead fabric bring-up, validation, and troubleshooting.
  • Build automation templates and tooling so each cluster starts from a known-good baseline.

Andromeda is a market and infrastructure platform to buy, sell, and operate compute for AI, enabling fragmented compute to flow through one platform. The company has a high talent density, high trust team and offers comprehensive benefits.

  • Support frontline NOC and resolve network faults in production infrastructure.
  • Lead incident management and drive prevention and automation efforts.
  • Create and maintain documentation, standards, and runbooks for network infrastructure.

Highrise AI is a GPU-native infrastructure platform designed for modern AI workloads. As a subsidiary of Hut 8 Corp., it offers a growing global team and a meritocratic, collaborative culture.

$270,400–$401,600/yr
US

  • Design and build data center network fabrics for large GPU clusters.
  • Automate network provisioning, configuration, and change management in code.
  • Own BGP, VLANs, VRFs, EVPN/VXLAN, and tune GPU networking.

A late-stage AI cloud company backed by NVIDIA with a valuation in the billions, running AI compute from data centers globally. They build dedicated, isolated GPU clusters for customers and maintain a fast-paced, engineering-driven culture.

US Unlimited PTO

  • Operate and configure Juniper PTX routers and Smartoptics optical line systems for edge and backbone connectivity.
  • Manage internet transit, peering, and external connectivity using BGP, IPv4/IPv6, and automation with Python and Ansible.
  • Troubleshoot packet and optical layer issues, perform capacity planning, and maintain network documentation and runbooks.

Lightning AI builds an end-to-end platform for developing, training, and deploying AI systems based on PyTorch Lightning. Founded in 2019, the company operates globally with offices in major cities and is backed by top venture capital firms.

United States

  • Serve as a subject matter expert for NVIDIA Networking technologies including Ethernet and InfiniBand-based data center and AI fabric solutions.
  • Participate in the design and deployment of complex solutions, create written deliverables, and conduct client workshops.
  • Communicate architecture strategy, implementation approach, and operational considerations in clear, concise terms to client stakeholders.

AHEAD builds platforms for digital business, weaving together cloud, automation, analytics, and software delivery to enable digital transformation. The company prioritizes a culture of belonging and values diverse perspectives, with a focus on empowering employees.

$270,400–$461,800/yr
US

  • Own networking across the managed Kubernetes platform, from cluster networking to data center fabric and GPU interconnect.
  • Automate tenant and node network lifecycles with Kubernetes-native tooling, including BGP, VLANs, overlays, and GPU networking.
  • Drive network telemetry, incident response, and security policy through GitOps and segmentation standards.

A late-stage AI cloud company backed by NVIDIA with a multi-billion-dollar valuation, running AI compute out of data centers worldwide. It is building a managed Kubernetes platform that turns fleets of bare-metal GPU compute into dedicated, isolated clusters for AI training and inference.

Global

  • Lead deployment and integration of ITVA across complex enterprise customer environments, including cloud, on-premises, and air-gapped infrastructure.
  • Implement AI-powered network operations, troubleshooting, visibility, and automation workflows.
  • Collaborate with ITVA's software engineers to translate customer challenges into product requirements.

ITVA builds an AI-native platform that combines network visibility, automation, troubleshooting, cybersecurity, and IT operations. ITVA is an engineering-driven company deploying into complex enterprise environments, focused on real-world impact.

Global 5w PTO

  • Define the technical direction and architecture for a rapidly expanding GPU cluster platform.
  • Lead and mentor a team of senior infrastructure engineers while staying hands-on with complex systems.
  • Establish security strategies and operational practices to ensure reliable, efficient AI experiments.

This organization powers cutting-edge AI research through large-scale GPU cluster infrastructure. It fosters a collaborative, fast-moving environment where engineers work closely with researchers.

Canada

  • Drive product strategy and execution for the vMetal bare-metal infrastructure platform, focusing on compute and networking.
  • Own the networking layer, including tenant network provisioning across InfiniBand and high-speed Ethernet fabrics.
  • Translate complex infrastructure requirements into detailed product specifications and engineering plans.

They build a bare-metal infrastructure platform designed for AI cloud providers and hyperscalers. They are a fast-growing, remote-first technology company with a supportive, distributed team culture.

US Unlimited PTO

  • Run cluster builds on site from hardware delivery to ready-for-service.
  • Direct integrators and contractors daily against the build plan.
  • Hand over validated clusters to operations with documentation and runbooks.

Andromeda is a market and infrastructure platform for buying, selling, and operating compute, enabling fragmented GPU capacity to flow through one platform. They are a high-talent-density team with comprehensive benefits, shaping the AI infrastructure market.