AI Infrastructure Engineer

AI Infrastructure Engineer
Responsibilities
1. NVIDIA GPU & Hardware Infrastructure Deployment

• Hardware Provisioning: Rack, stack, configure, and maintain high-performance bare-metal GPU servers (e.g., NVIDIA H200, B300 or equivalent Supermicro/Dell/HGX architectures).
• AI Software Stack: Install, update, and optimize NVIDIA Drivers, CUDA Toolkit, cuDNN, and NVIDIA Container Toolkit on physical host machines.
• Containerization & Orchestration: Manage GPU-accelerated environments using Docker, including configuring GPU partitioning (MIG/vGPU) for optimal resource allocation.

2. Network & Systems Engineering

• High-Performance Networks: Configure and optimize InfiniBand (IB) switches and RoCE (RDMA over Converged Ethernet) to ensure ultra-low latency and maximum throughput for multi-GPU training workloads.
• Core Infrastructure: Manage enterprise firewalls, core switches, VLANs, and local network routing to ensure high security and stability of the data center network.
• Linux Administration: Oversee Linux server administration (Ubuntu, RHEL, or Rocky Linux), including automated OS provisioning and local storage clusters.

3. Metering & Billing System Integration

• Resource Metering: Implement and configure telemetry tools to accurately monitor and log GPU time, CPU utilization, storage usage, and network traffic.
• Billing System Management: Maintain and integrate usage-based billing/metering engines to track infrastructure costs or client usage.
• Automation: Write robust scripts (Python, Go, or Bash) to link data center resource telemetry with the billing platform for precise invoicing and automated usage reporting.

Qualifications & Skills
Required Qualifications:

• Experience: 3-5+ years of experience in Network Engineering, Linux Systems Administration, or DevOps, with hands-on experience in GPU infrastructure deployment.
• Linux & Automation: Expert-level knowledge of Linux environments and infrastructure-as-code/automation tools (Ansible, Terraform, or SaltStack).
• NVIDIA Ecosystem: Deep technical understanding of the NVIDIA AI Enterprise stack (CUDA, NCCL, NVLink).
• Billing/Metering Awareness: Practical experience working with usage-based tracking, billing APIs, or internal chargeback tools.

Pay: $80,000.00 - $120,000.00 per year

Benefits:

• Dental insurance
• Health insurance
• Paid time off
• Vision insurance

Work Location: In person