> Markdown version of [/jobs/ext/4685492-staff-cluster-infrastructure-engineer](https://www.wearedevelopers.com/jobs/ext/4685492-staff-cluster-infrastructure-engineer). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Staff Cluster Infrastructure Engineer

- **Company:** ATOM, INC.
- **Location:** San Francisco, CA, United States
- **Experience:** Expert
- **Salary:** $224,000 - $284,000
- **Contract:** Permanent contract
- **Skills:** Artificial Intelligence, Python (Programming Language), Linux System Administration, Software Engineering, Systems Integration, Graphics Processing Unit (GPU), Cloudformation, Kubernetes, Infrastructure Automation Frameworks, Bare Metal, Terraform
- **Published:** November 12, 2026
- **Apply:** https://www.careerbuilder.com/job-details/staff-cluster-infrastructure-engineer-san-francisco-ca--4ec1afe5-e5d3-4707-aab3-7d4b6d852aeb

## About the Role

Mineral/Metal Mining, Artificial Intelligence (AI), Real Estate, Manufacturing, Hardware-Software Integration, GPU (Graphics Processing Unit), Simulation, Automation, Identify Issues, Problem Solving Skills, * 6+ years experience operating GPU compute on Kubernetes (or similar orchestration), with the judgment to scale it as demand grows.

 * Strong programming and scripting skills in Python, Go, or similar.
 * Familiarity with Infrastructure-as-Code tools such as Terraform or CloudFormation.
 * Comfort with bare-metal Linux environments, GPU hardware, and networking.
 * A bias toward automation, reliability, and operating critical systems well.

## Description

We're seeking a Staff Cluster Infrastructure Engineer to join our founding team who will own the GPU compute fabric that trains our foundation models - optimizing the machines we have today, automating how we manage them, and laying the groundwork to scale as we grow.

 * Manage and automate our GPU training clusters, including provisioning, bootstrapping, and lifecycle management.
 * Automate bare-metal bring-up so new machines come online quickly and reliably as we add capacity.
 * Build software abstractions that present a clean, unified interface to our training and simulation workloads.
 * Work at the hardware/software boundary, where speed and reliability are critical, continuously raising the bar for automation and uptime.
 * Run day-to-day operations: diagnose and resolve issues quickly when systems are under pressure.
 * Design our infrastructure to scale smoothly as we grow from a smaller cluster of machines toward a larger fleet.