Field Engineer

About the role

Launched in 2019, Latitude.sh provides automated, global Bare Metal infrastructure with cloud-like flexibility. Recently acquired by Megaport, the company is integrating high-performance compute with world-class connectivity to power the next generation of AI and enterprise workloads at scale.

The Datacenter Operations team keeps the physical infrastructure behind the bare-metal cloud running: servers, racks, power, cabling and structured connectivity across a global data center footprint. A fast-growing part of that footprint is high-density GPU infrastructure — multi-GPU AI servers (8× SXM-class NVIDIA platforms) drawing well above typical enterprise rack densities, with demanding power, cooling and cabling requirements. These systems run production AI workloads for customers, so precision, uptime and hardware discipline are non-negotiable.

This is a hands-on Field Engineer role acting directly on that physical infrastructure. You will deploy, maintain, troubleshoot and repair GPU servers and their supporting infrastructure, travel between data centers in your region, and act as the on-site focal point for new deployments, including overseeing remote hands teams at partner facilities. The position is hired at a senior level; mid-level candidates will be considered when they bring solid, demonstrable GPU server hardware experience.

Responsibilities

GPU server maintenance & break-fix

• Diagnose and repair hardware failures on multi-GPU servers (SXM baseboards, GPU trays, NVLink components, PCIe risers, NICs, DIMMs, NVMe drives, PSUs, fans)
• Perform component-level replacement following vendor procedures and ESD-safe handling practices
• Run and interpret hardware diagnostics (BMC/IPMI logs, vendor diagnostic tools, GPU health checks such as nvidia-smi, DCGM and field diagnostics) to isolate faults before intervention
• Execute firmware updates (BIOS, BMC, GPU, NIC) as part of maintenance windows and break-fix workflows
• Manage the full RMA cycle: fault evidence collection, vendor ticket handling, part logistics, replacement and validation
• Validate repaired systems back to production standards (burn-in, stress tests, link and thermal checks) before customer handover

Deployments, racking & cabling

• Rack, stack and cable GPU servers and supporting equipment (switches, PDUs), respecting weight, airflow and power-density constraints of high-density racks
• Perform structured copper and fiber cabling: run, terminate, label, test and certify links (including high-speed interconnects such as 400G/NDR-class fabrics)
• Verify power distribution readiness (PDU capacity, circuit balancing, redundancy) before energizing new GPU racks
• Execute server installations and decommissions end to end, including physical audit and inventory updates
• Support the Deploy & Implementation team by confirming switch, PDU and cross-connect availability prior to activation

New deployment oversight & remote hands coordination

• Act as the on-site focal point for new GPU deployments in your region
• Plan, brief and oversee remote hands teams at partner facilities: define work scopes, validate execution quality and sign off on completed work
• Ensure deployment work follows Latitude.sh standards for cabling, labeling, documentation and safety
• Report deployment progress, blockers and quality issues to Datacenter Operations leadership and coordinate corrective actions

Operations & documentation

• Respond to physical infrastructure incidents within agreed SLAs, including on-call escalations
• Maintain accurate, up-to-date documentation: rack elevations, cabling maps, asset inventory, maintenance records and runbooks
• Perform preventive maintenance and inspection routines on servers and rack-level infrastructure
• Follow and improve change management procedures for all physical interventions in production environments
• Collaborate with Network Operations, Deploy & Implementation and Electrical teams on cross-functional activities

Requirements

• Advanced English communication skills (C1 level or higher), both written and spoken
• Proven hands-on experience with datacenter server hardware: installation, troubleshooting, component replacement and validation; demonstrated skill valued over years of experience
• Solid knowledge of GPU server platforms and architecture, including:
  • SXM/HGX baseboard systems (e.g. NVIDIA H100/B-series class platforms) and PCIe GPU servers
  • GPU interconnects: NVLink/NVSwitch, PCIe topologies
  • High-power server design: multi-PSU configurations, power budgeting (10–15 kW class systems), airflow and thermal behavior
• Experience with out-of-band management (BMC, IPMI, Redfish, vendor consoles such as iDRAC/Supermicro IPMI) for diagnostics and remote troubleshooting
• Competence in structured cabling: copper and fiber standards, connector types (LC, MPO/MTP, OSFP/QSFP transceivers, DACs/AOCs), testing and certification practices
• Understanding of datacenter power fundamentals: PDU types, single vs. three-phase circuits, redundancy models (A/B feeds), and safe work practices around energized equipment
• Ability to read and produce technical documentation: rack elevations, cabling schedules, work instructions and incident reports
• Strong troubleshooting methodology: able to isolate faults systematically and document root cause
• Autonomy and ownership: comfortable being the senior technical presence on site, making sound decisions under pressure and knowing when to escalate
• Availability for regional travel between data centers and for on-call rotations outside business hours
• Physical capability for datacenter work: lifting and handling heavy equipment (GPU servers routinely exceed 100 kg per chassis, handled with lifts and team procedures), working in hot/cold aisles

Qualifications

• Experience with direct liquid cooling (DLC) systems for GPU servers: cold plates, quick disconnects, CDUs, manifolds, leak detection and coolant handling procedures
• Hands-on experience with high-density AI platforms from Supermicro, Dell, NVIDIA (HGX/DGX-class) or similar OEMs
• Familiarity with InfiniBand/NDR fabrics and their physical layer (OSFP optics, splitter cables, fabric cabling patterns)
• Certifications such as CompTIA Server+, NVIDIA-Certified (hardware/AI infrastructure tracks), BICSI Installer/Technician, or vendor hardware certifications (Dell, Supermicro, HPE)
• Experience working in or coordinating with colocation/partner facilities and their remote hands processes
• Basic Linux command-line skills for hardware validation and diagnostics
• Background in bare-metal, hyperscale, HPC or AI infrastructure environments

Schedule

• Reports into the Datacenter Operations team
• Business hours with on-call rotation for after-hours physical infrastructure incidents
• Regional scope: based near one of our data center locations, with regular travel between nearby facilities for deploy