GPU Infrastructure Software Engineer I

crusoe; San Francisco (CA); On-site; USD 117,000 - 135,000; Full time

Benefits offered by this job

Health insurance

401(k) with match

Stock options/RSUs

Parental leave

Life insurance and disability

Commuter benefit

Job summary

Crusoe is seeking a Software Engineer to join the Data Center Infrastructure Engineering team. You will develop software to manage a large GPU server fleet and the data centers they inhabit, focusing on diagnostics, observability, automation, and repair tooling for high-performance GPU compute clusters.

The role requires hands-on problem-solving, independence, and a drive to scale Crusoe’s GPU fleet while ensuring reliability and performance for customers and partners.

Qualifications

• 1 year of software engineering experience.
• Ability to identify a problem, rapidly develop a scalable solution and ship it.
• Strong analytical and problem-solving skills.
• Excellent communication and collaboration skills.

Responsibilities

• Develop deep-level diagnostics for GPU racks and high-density compute systems.
• Develop automation tooling for GPU platforms (NVIDIA A100, H200, GB200, B200; AMD 350X/355X).
• Create automation and AI agents for component-level diagnosis and remediation.
• Develop tooling for post-repair validation and testing (burn-in, PyTorch, NCCL).
• Own deployment, monitoring, and support of tooling to maximize GPU fleet availability.

Skills

Distributed systems

Reliability

Cloud platforms

Go

Python

Java

Rust

Team collaboration

Tools

Kubernetes

Temporal

IaC

GCP

Job description

Crusoe is seeking a Software Engineer to join the Data Center Infrastructure Engineering team. You will develop software to manage a large GPU server fleet and the data centers they inhabit, focusing on diagnostics, observability, automation, and repair tooling for high-performance GPU compute clusters.

The role requires hands-on problem-solving, independence, and a drive to scale Crusoe’s GPU fleet while ensuring reliability and performance for customers and partners.