Senior Kubernetes Engineer — GPU HPC & AI Platforms

NorthMark Compute & Cloud; Dallas (TX); On-site; USD 140,000 - 210,000; Full time

Job summary

NorthMark Compute & Cloud in Dallas seeks a Senior Kubernetes Engineer to design, implement, and optimise GPU-accelerated container platforms at scale, spanning hybrid or on-prem environments. The role requires deep NVIDIA and Kubernetes expertise, including GPU Operator, device plugins, MIG, and advanced scheduling.

You will drive multi-tenant, high-throughput workloads for AI/ML, HPC and LLM training, while enhancing observability and security with GitOps-driven delivery.

Qualifications

• Experience operating Kubernetes in production environments with NVIDIA GPUs (Operator, device plugin, MIG).
• Proficiency in Go or Python for operator/controllers development.
• Deep understanding of Kubernetes internals: CRDs, RBAC, custom controllers and scheduler extensions.
• Experience with GPU-intensive workloads (LLMs, HPC, ML pipelines).
• Hands-on with Helm, Kustomize and GitOps workflows.

Responsibilities

• Architect and operate GPU-optimised Kubernetes clusters using NVIDIA Operator, DCGM, and network tools.
• Develop and maintain custom Kubernetes operators and controllers to automate infra services.
• Integrate NVIDIA device plugins, MIG, and GPU sharing into scheduling layer.
• Tune scheduling with kube-scheduler plugins, Slurm, Volcano for high throughput.

Skills

Kubernetes in prod

Go or Python

RBAC & security

GPU workloads

GitOps workflows

Education

Bachelor's degree in CS or related

Tools

NVIDIA GPU Operator

DCGM

Helm

Kustomize

ArgoCD

FluxCD

Terraform

Prometheus

Grafana

Job description

NorthMark Compute & Cloud in Dallas seeks a Senior Kubernetes Engineer to design, implement, and optimise GPU-accelerated container platforms at scale, spanning hybrid or on-prem environments. The role requires deep NVIDIA and Kubernetes expertise, including GPU Operator, device plugins, MIG, and advanced scheduling.

You will drive multi-tenant, high-throughput workloads for AI/ML, HPC and LLM training, while enhancing observability and security with GitOps-driven delivery.