Remote GPU Infra SRE for Large-Scale AI Training

andromeda hill; San Francisco (CA); Hybrid; USD 180,000 - 260,000; Full time

Job summary

Andromeda is seeking a Senior Site Reliability Engineer to design, operate, and optimize large-scale GPU infrastructure used for distributed training and inference. You will work directly with customers to push the limits of AI systems and ensure reliable capacity across multiple regions.

You will own GPU cluster architecture, performance, and reliability engineering, with emphasis on network fabrics, observability, automation, and incident leadership in production environments.

Qualifications

• Deep, hands-on experience operating large-scale GPU clusters.
• Experience with GPU memory hierarchies, ECC behavior, thermal throttling, and hardware failure modes.

Responsibilities

• GPU Cluster Architecture: design and evolve multi-provider, multi-region GPU clusters.
• Customer Technical Partnership: onboard and troubleshoot for customers running large-scale training workloads.
• Reliability & Performance Engineering: define SLOs and budgets for GPU infrastructure.
• Networking & Fabric Health: ensure health of InfiniBand, RoCE, NVLink fabrics.

Skills

GPU systems

High-performance networking

Distributed training

Linux internals

Kubernetes

Automation

Observability

Incident management

Tools

CUDA toolkit

NCCL

InfiniBand

Job description

Andromeda is seeking a Senior Site Reliability Engineer to design, operate, and optimize large-scale GPU infrastructure used for distributed training and inference. You will work directly with customers to push the limits of AI systems and ensure reliable capacity across multiple regions.

You will own GPU cluster architecture, performance, and reliability engineering, with emphasis on network fabrics, observability, automation, and incident leadership in production environments.