Senior GPU Interconnect & Platform Reliability Engineer

Firmus Technologies; City of Melbourne; On-site; AUD 180,000 - 240,000; Full time

Job summary

Firmus Technologies in Australia leads AI infrastructure across APAC. The Senior Platform Reliability Engineer, Fabric and Interconnect, owns the reliability of the GPU interconnects this estate runs on, the high-performance network fabrics, and DPU-based host networking that binds compute to the platform.

This is a hands-on senior role with automation as a first-class component: you will build and maintain guarded automation and remediation tooling, reproduce faults with vendors, and drive

Qualifications

• 8+ years in production networking or interconnect infra in 24/7 environments.
• Deep knowledge of GPU interconnect fabrics and topology, fault diagnosis.
• Experience with InfiniBand, RoCE, or NVIDIA Spectrum-X fabrics.
• DPU/SmartNIC host networking, offload paths and firmware coordination.
• Automation and IaC practices with peer review and progressive rollout.
• Scripting for automation using Python, Go, or Bash.
• Ability to act as senior escalation point and produce runbooks.
• Strong technical judgment and clear communication for complex faults.

Responsibilities

• Responsible for reliable operation, automation and continuous improvement of GPU interconnect and network fabrics (NVLink, NVSwitch, InfiniBand, Spectrum-X).
• Build and maintain guarded automation and remediation tooling for fabric faults and AI clusters.
• Diagnose and tune performance across interconnect stack from application to link level.
• Operate DPU-based host networking across the fleet, including offload path configuration.
• Execute firmware upgrade waves, fabric expansions and capacity changes with rollback and verification.
• Provide deep expertise for fabric faults and drive permanent fixes with engineering collaboration.
• Lead vendor escalations at engineering level and mentor frontline engineers.
• Lead technical recovery during major fabric incidents and document runbooks and procedures.

Skills

High-performance networking

GPU interconnect

InfiniBand/RoCE

DPU/SmartNIC

Firmware upgrades

Automation / IaC

Scripting (Python/Go/Bash)

Incident response / Runbooks

Communication & escalation

Job description

Firmus Technologies in Australia leads AI infrastructure across APAC. The Senior Platform Reliability Engineer, Fabric and Interconnect, owns the reliability of the GPU interconnects this estate runs on, the high-performance network fabrics, and DPU-based host networking that binds compute to the platform.

This is a hands-on senior role with automation as a first-class component: you will build and maintain guarded automation and remediation tooling, reproduce faults with vendors, and drive