AI Infra Network Engineer for High-Speed GPU Clusters

SpaceX; Hawthorne (CA); On-site; USD 125,000 - 195,000; Full time

Job summary

SpaceX's Starshield division seeks a Network Engineer to own the data center network design, deployment, and security for critical national security missions. You will build high-speed copper and optical interconnects for AI clusters and drive automation for rapid deployments.

You will collaborate with ML teams to translate workload patterns into topology, monitor performance, and improve reliability. A degree or 4+ years in networking is required, plus Layer-1 expertise and clearance

Qualifications

• Bachelor's degree or 4+ years networking experience in lieu of degree.
• 1+ year of enterprise/networking experience, including GPU networks.
• Layer-1 experience with fiber, DWDM, or related tech.

Responsibilities

• Manage network infrastructure for top secret data centers.
• Design high-speed copper and optical interconnects for AI clusters.
• Collaborate with ML teams to map workload patterns to topology.
• Develop tools for rapid deployments (ZTP/provisioning).
• Improve performance, reliability, and capacity planning.
• Lead network integration and troubleshooting efforts.
• Coordinate with hardware, software, and network engineers.

Skills

Layer-1 experience

Networking hardware deployment

High-speed interconnects

Network security

ML collaboration

Automation tooling

Clearance eligibility

Education

Bachelor's degree in technical or engineering discipline

4+ years networking experience

Master's degree (preferred)

Tools

Cisco certification

Juniper certification

Arista certification

ZTP tooling

Job description

SpaceX's Starshield division seeks a Network Engineer to own the data center network design, deployment, and security for critical national security missions. You will build high-speed copper and optical interconnects for AI clusters and drive automation for rapid deployments.

You will collaborate with ML teams to translate workload patterns into topology, monitor performance, and improve reliability. A degree or 4+ years in networking is required, plus Layer-1 expertise and clearance