AI Factory Network Architect — Remote-First & Flexible

Speedrun Talent Network; United States; Remote; USD 260,000 - 360,000; Full time

Benefits offered by this job

Generous paid time off

Stock options

Flexible hours

Family leave

Company retreats

Job summary

Runware is building the AI Factory network across US and Europe. You will own the architecture from server NVLink domains to inter-site backbones, lead hands-on fabric bring-up, and grow the team that operates tens of thousands of ports.

This is a hands-on leadership role with global collaboration, extensive travel, and responsibility for latency, throughput, and reliability in a cutting-edge HPC/AI environment.

Qualifications

• 15+ years designing, building, and operating high-performance networks for HPC/AI scales.
• Hands-on InfiniBand, fabric design, routing, and congestion control expertise.
• Knowledge of NVLink/NVSwitch in multi-GPU, multi-node setups.
• Deep optical networking experience for inter-site links and 400G/800G optics.
• RDMA across InfiniBand and RoCE, GPUDirect and NCCL integration.
• Proven leadership of network teams through greenfield builds and incidents.
• Automation incl. Python/Go, IaC, telemetry pipelines, and monitoring.
• Ability to translate designs into actionable site-level documentation.
• Willingness to travel to pod sites in US and Europe.

Responsibilities

• Own the end-to-end network architecture for the AI Factory across pods and backbone.
• Design the optical layer, link budgets, DWDM, and reliability targets.
• Define a validated reference network design for new pods to join quickly.
• Build and bring up fabrics, including routing, QoS, and RDMA tuning.
• Collaborate with inference engine teams to align traffic patterns with topology.
• Lead capacity planning, vendor evaluations, and roadmap alignment with partners.
• Set operating standards for telemetry, health monitoring, and fault isolation.
• Hire and develop a network engineering team as the footprint grows.

Skills

InfiniBand design

NVLink/NVSwitch

Optical networking

RDMA tuning

Python/Go

Infrastructure as code

Telemet ry pipelines

System bring-up

Leadership

Cross-site collaboration

Technical writing

Travel willingness

Tools

NVIDIA Spectrum-X / Quantum-X800 / Blackwell NVLink

Co-packaged optics / photonic switching

SONiC / OpenFabrics / UEC

Job description

Runware is building the AI Factory network across US and Europe. You will own the architecture from server NVLink domains to inter-site backbones, lead hands-on fabric bring-up, and grow the team that operates tens of thousands of ports.

This is a hands-on leadership role with global collaboration, extensive travel, and responsibility for latency, throughput, and reliability in a cutting-edge HPC/AI environment.