HPC AI Infrastructure Architect

MRE Consulting; Houston (TX); On-site; USD 120,000 - 180,000; Full time

Benefits offered by this job

Competitive salary

Comprehensive benefits

Professional development support

Job summary

MRE Consulting is seeking a skilled HPC AI Systems Administrator to architect and optimize a secure on-prem AI compute environment. You will enable the Development Team to train and deploy production ML models while enforcing governance and data privacy policies.

The role covers deployment, bare-metal config, GPU driver management, and containerized workloads. Collaboration with Dev teams and vendors is essential to maintain peak performance and security.

Qualifications

• 3+ years of systems administration in Linux-based HPC or enterprise GPU environments.
• Experience with modern enterprise GPU hardware and data center contexts.
• Deep expertise in Linux, container tech, and cluster resource management.

Responsibilities

• Lead deployment, bare-metal configuration, and optimization of on-prem HPC cluster and multi-GPU systems.
• Manage AI software stack, Linux environments, GPU drivers, CUDA/NCCL, and container platforms.
• Implement workload scheduling/orchestration (Kubernetes, Slurm) for efficient resource allocation.
• Establish telemetry/monitoring dashboards for hardware utilization and performance.
• Ensure security/governance, data-at-rest/in-transit baselines, and compliance.
• Coordinate with vendors and integrators for updates and platform maintenance.

Skills

Linux admin

HPC infrastructure

Containerization

Cluster scheduling

Networking

Security governance

Education

Bachelor's degree (CS/CE)

Tools

Docker

Kubernetes

Slurm

Apptainer/Singularity

CUDA drivers

Job description

MRE Consulting is seeking a skilled HPC AI Systems Administrator to architect and optimize a secure on-prem AI compute environment. You will enable the Development Team to train and deploy production ML models while enforcing governance and data privacy policies.

The role covers deployment, bare-metal config, GPU driver management, and containerized workloads. Collaboration with Dev teams and vendors is essential to maintain peak performance and security.