Senior GPU Infra SRE — Remote, Low-Level Linux & Scale

Luma; Redwood City (CA); Remote; USD 180,000 - 240,000; Full time

Job summary

Luma, based in California, seeks a Senior SRE to own the GPU infrastructure that powers our research and product workloads across on-prem, AWS, and OCI. You will keep training and inference clusters reliable and fast, collaborating on scale across multiple environments.

This hands-on role is Linux-focused and near the metal—debugging GPU, networking, and kernel-level issues, including direct work with NVIDIA. If you want a narrowly scoped ops job, this isn’t it.

Qualifications

• 5+ years as an SRE, production, or infra engineer in a fast-paced, large-scale environment.
• Deep Linux expertise, containerized systems, and low-level performance debugging.
• Experience with AWS or OCI and high-performance networking (InfiniBand/RDMA).
• Security and compliance familiarity (SOC 2 / ISO).

Responsibilities

• Take end-to-end ownership of production GPU clusters for training and inference across AWS and OCI.
• Join critical re-architecture sessions to redesign systems for higher efficiency and scale.
• Tune Linux performance deeply, at the OS and kernel level.
• Build automation in Python, Go, or Bash to manage, monitor, and self-heal infrastructure without heavy toil.
• Serve as the final escalation for the hardest GPU, networking, and system failures, working with vendors like NVIDIA.
• Help achieve and maintain security certifications (SOC 2 Type 1 & 2, ISO) with strong infrastructure security practices.

Skills

Linux kernel debugging

GPU infrastructure

Terraform

Airflow

Ray

AWS

OCI

InfiniBand/RDMA

SOC 2 / ISO security

Kubernetes

GPU tooling (DCGM/ROCm)

Large-scale GPU clusters

Kubernetes / Ray orchestration

Data pipelines & infra

Tools

Terraform

Airflow

Ray

Kubernetes

Job description

Luma, based in California, seeks a Senior SRE to own the GPU infrastructure that powers our research and product workloads across on-prem, AWS, and OCI. You will keep training and inference clusters reliable and fast, collaborating on scale across multiple environments.

This hands-on role is Linux-focused and near the metal—debugging GPU, networking, and kernel-level issues, including direct work with NVIDIA. If you want a narrowly scoped ops job, this isn’t it.