Senior AI Systems Administrator

Overview

In this role you will manage multi-rack GPU clusters used for AI training and inference, from hardware to orchestration. You will work with researchers and engineers to improve system performance and reliability. The position sits at the intersection of HPC, networking and DevOps in a mission-driven AI organisation. You’ll be empowered to shape how cutting-edge compute is provisioned and operated, with a strong focus on collaboration and impact.

Pay / Benefits

• pension
• benefits
• professional development
• city-centre office near Cambridge station

Responsibilities

• Deploy, tune and manage GPU servers across multiple racks
• Build and operate high-performance networks (InfiniBand, RDMA, RoCE) and related switches/routers/firewalls
• Administer Linux across multiple distributions, including hypervisors and virtual machines
• Manage infrastructure as code using Git and Terraform/OpenTofu
• Collaborate with researchers and engineers to continuously improve system performance

Key requirements

• Significant experience as Systems Administrator, HPC Engineer, MLOps Engineer or similar
• Hands-on GPU server deployment and tuning
• Strong InfiniBand/RDMA/RoCE knowledge and solid network design skills
• Advanced Linux, virtualization, Git and infrastructure-as-code experience
• Collaborative approach to working with system users

• collaborative
• user-focused
• problem-solving
• GPU server deployment and tuning
• InfiniBand, RDMA, RoCE
• Linux (Ubuntu, RHEL) and virtualization

…