Contract Site Reliability Engineer — AI Accelerator Infrastructure

About the role

This is a one-year contract with potential for a full-time conversion. This is an ownership role, not a ticket executor role. You operate independently, document what you build, and escalate with precision when something needs engineering attention.

Responsibilities

• Physical Infrastructure & Hardware Bring-Up
• Rack, stack, cable, and decommission servers, PDUs, network gear, and storage in on-premises labs and colocation facilities.
• Execute hardware bring-up for d-Matrix AI-accelerated systems and validation test benches, including BIOS configuration, firmware validation, and OS installation (Linux primary).
• Replace and upgrade components: PCIe cards, NICs, storage, memory, and accelerator hardware across multiple server generations.
• Maintain lab spaces in organized, ESD-compliant, and audit-ready condition at all times.
• Asset Management & Inventory
• Own accurate asset tracking and inventory records using DCIM tools (NetBox, Jira, or equivalent); every piece of hardware is accounted for with the current configuration state.
• Cook up equipment moves between lab, staging, and colocation; manage shipping, receiving, and RMA workflows.
• Track hardware lifecycle status: warranty, EOL, refresh schedules, and spare parts inventory.
• Troubleshooting & Incident Support
• Diagnose and resolve hardware failures, power and thermal issues, and network connectivity problems — escalating to SRE engineers with clear documentation and reproduction steps.
• Serve as the first physical responder for infrastructure incidents requiring hands-on intervention in lab or colo environments.
• Documentation & Collaboration
• Write and maintain SOPs, rack diagrams, and cabling guides clear enough for a new technician to execute without shadowing.
• Communicate issues and status across hardware, software, SRE, and silicon validation teams professionally and with precision.
• Support parallel workstreams as silicon development programs evolve; operate independently when SREs are heads-down on engineering work.

Requirements

• 3+ years in data center operations, hardware validation, or systems technician roles, hands-on with real servers.
• Proven rack-and-stack experience: physical server installation, PDU and patch panel cabling, rack power planning, and cable management to professional standards.
• Hardware bring-up experience: BIOS/UEFI configuration, firmware updates, component replacement, and OS installation across Linux distributions.
• Linux command-line proficiency: enough to run diagnostics, inspect logs, and execute runbooks independently.
• Asset management discipline: experience with DCIM or inventory tools and a track record of accurate, up-to-date records.
• Strong written communication: you document what you do, escalate with context, and write SOPs others can follow.
• Comfortable operating independently in a fast-moving startup environment.

Preferred Qualifications

• Prior colocation data center experience (Equinix, CoreSite, Digital Realty, or similar).
• Experience with AI/ML or GPU hardware: NVIDIA, AMD, or Intel accelerator cards in a lab or production environment.
• Basic Ansible exposure — able to run playbooks and understand what they do.
• Familiarity with high-speed interconnects: InfiniBand, RoCE, SFP/QSFP optics.
• Python or Bash scripting for operational task automation.