HPC Platform Engineer (AI)

TL;DR

HPC Platform Engineer (AI): Owning the on-prem GPU and HPC platform lifecycle with an accent on provisioning, GPU orchestration, and cross-domain integration. Focus on building a dependable, high-performance service by stitching together scheduling, networking, storage, and compute to support real-world AI training and inference workloads.

Company

hirify.global is a technology company focused on building and maintaining high-performance computing and GPU infrastructure.

What you will do

• Manage bare-metal provisioning, OS imaging, and firmware/driver lifecycles.
• Orchestrate GPU workloads using Kubernetes with the NVIDIA GPU Operator or Slurm.
• Integrate scheduling, networking, storage, and compute layers into a coherent platform.
• Ensure platform availability and predictability for compute-intensive AI workloads.
• Drive operational excellence through automation, capacity planning, and incident response.
• Collaborate with network and storage engineers to optimize fabric design and I/O patterns.

Requirements

• 5+ years of experience operating production Linux infrastructure at scale in HPC or GPU environments.
• Strong proficiency in Linux fundamentals, including kernel/driver troubleshooting and performance debugging.
• Hands-on experience with bare-metal automation (PXE/iPXE, MAAS, Redfish) and configuration management (Ansible, Terraform).
• Deep knowledge of GPU operations, including CUDA, NVIDIA Container Toolkit, and DCGM telemetry.
• Experience with HPC schedulers like Slurm or Kubernetes-based GPU orchestration.
• Fluent English required for documentation and cross-team coordination.

Nice to have

• Experience with multi-tenant GPU-as-a-Service environments.
• Familiarity with hybrid Slurm and Kubernetes workflows.
• Low-level diagnostics skills (NUMA, PCIe topology, IRQ affinity).
• Contributions to open-source HPC or Kubernetes tooling.

Culture & Benefits

• Opportunity to work on high-performance, large-scale GPU infrastructure.
• Focus on operational excellence and reducing technical toil through automation.
• Collaborative environment working closely with specialized network and storage engineering teams.
• Emphasis on disciplined incident response and measurable platform improvements.