Remote-First GPU Infrastructure Engineer – On-Prem Kubernetes

Orcrist Technologies GmbH; United States; Remote; USD 102,000 - 171,000; Full time

Benefits offered by this job

Remote-first

Home office budget

Great equipment

30 days vacation

Job summary

Orcrist Technologies GmbH is hiring an Infrastructure Engineer – Platform to own bare-metal GPU servers, OS, networking, and storage across on-prem and air-gapped sites. You will manage the NVIDIA stack and contribute to Kubernetes on bare-metal, while partnering with SRE/ML teams for efficient on-prem inference.

Responsibilities include sizing, racking, commissioning environments, and enabling secure, scalable deployments in a remote-first setting focusing on Germany.

Qualifications

• 5+ years in bare-metal, HPC/GPU, or data-center infra.
• Strong bare-metal Linux (RHEL/Rocky/Ubuntu) and networking basics.
• Experience with NVIDIA GPU stack and on-prem inference.
• Willing to travel to customer sites.
• Documentation-focused and calm handling hardware incidents.
• Eligible to work in Germany.

Responsibilities

• Design, size, provision, and operate bare-metal GPU server fleets across on-prem and air-gapped environments.
• Own the NVIDIA GPU stack end to end: drivers, CUDA, GPU Operator, MIG, and DCGM.
• Build the bare-metal substrate Kubernetes runs on: node lifecycle, container runtime, plugins, tuning.
• Engineer data-center networking and resilient storage sized to scale with encryption at rest.
• Partner with ML/MLOps on on-prem inference serving and performance tuning.
• Plan and run on-site build-outs: rack, power, cooling, capacity, runbooks.

Skills

Bare-metal infrastructure

Linux systems

NVIDIA GPU stack

On-prem / air-gapped

Documentation & calm under hardwareInc

Tools

CUDA

NVIDIA GPU Driver management

MAAS/Metal3

Ansible

Terraform/Pulumi

PXE/iPXE

Job description

Orcrist Technologies GmbH is hiring an Infrastructure Engineer – Platform to own bare-metal GPU servers, OS, networking, and storage across on-prem and air-gapped sites. You will manage the NVIDIA stack and contribute to Kubernetes on bare-metal, while partnering with SRE/ML teams for efficient on-prem inference.

Responsibilities include sizing, racking, commissioning environments, and enabling secure, scalable deployments in a remote-first setting focusing on Germany.