Senior GPU Compute Infra Engineer — Remote

Boundless Networks, Inc.; United States; Remote; USD 150,000 - 210,000; Full time

Benefits offered by this job

Salary + equity

Health, dental, vision

Flexible PTO

Remote-first with off-sites

Job summary

Boundless Networks, Inc. is seeking a Senior Infrastructure Engineer (GPU Compute) to design and operate a global, heterogeneous GPU fleet. You will optimize bare-metal and CUDA performance, orchestrate workloads across on-prem and cloud providers, and push efficiency while maintaining reliability.

You will own scheduling, resource allocation, and observability for a scalable inference substrate, with a bias for action and autonomy in a fast-moving AI compute team.

Qualifications

• 5+ years of infrastructure/DevOps experience managing large-scale production systems.
• Deep expertise in Kubernetes, Docker, and container orchestration at scale.
• Strong Linux systems administration skills.
• Proficiency with IaC tools Terraform, Ansible, and Pulumi.
• Experience with high-throughput, mission-critical systems.
• Scripting or programming in Python, Bash, TypeScript, or Go.
• Ability to navigate ambiguity with a bias for action.

Responsibilities

• GPU Fleet Orchestration: operate a multi-region GPU fleet across regions and providers.
• Compute Scheduling & Utilization: maximize GPU utilization across inference workloads.
• Bare-Metal & GPU Optimization: optimize PCIe, ReBAR, NUMA, CUDA tuning.
• Reliability, Access & Observability: build secure access, observability, zero-downtime rollouts.
• Cost Optimization: reduce $/GPU-hr via smart scheduling and placement.

Skills

Infra/DevOps

Kubernetes

Docker

Linux admin

Infrastructure as Code

Scripting

Python

TypeScript

Go

Tools

Terraform

Ansible

Pulumi

Job description

Boundless Networks, Inc. is seeking a Senior Infrastructure Engineer (GPU Compute) to design and operate a global, heterogeneous GPU fleet. You will optimize bare-metal and CUDA performance, orchestrate workloads across on-prem and cloud providers, and push efficiency while maintaining reliability.

You will own scheduling, resource allocation, and observability for a scalable inference substrate, with a bias for action and autonomy in a fast-moving AI compute team.