About the company
An early-stage deep-tech organisation is developing AI systems that learn from real-world experimentation to support materials discovery. By connecting machine learning with experimental research, it aims to help scientists investigate new possibilities and make scientific discovery faster and more effective.
The team brings together materials scientists, AI researchers and engineers in a fast-moving environment, building software and infrastructure around practical research needs. Collaboration, technical excellence and meaningful scientific impact are central to its approach. The organisation offers the opportunity to contribute to ambitious work at the intersection of AI and materials science, helping establish the foundations for research that could advance the field.
What the job entails
As Platform Engineer, you will build and own the compute infrastructure that enables researchers to train, run and deploy machine-learning models reliably. This is a foundational role: you will establish shared systems and practices that make research compute accessible, reproducible and efficient.
You will design and manage cloud infrastructure through infrastructure-as-code, and operate GPU environments for training and inference. This includes cluster configuration, scheduling and improving compute cost, utilisation and performance. You will develop orchestration and tooling for training workloads, maintain CI/CD pipelines, and support consistent environments from research through to production. You will also establish monitoring, alerting, security and incident-response practices, and create tools and documentation that help researchers self-serve. Working closely with researchers and engineers, you will focus on the compute platform rather than data engineering or model research itself.
Key requirements / What we look for
-
At least four years’ experience in platform engineering, DevOps or SRE, with hands-on infrastructure ownership.
-
Deep Kubernetes or equivalent orchestration experience, ideally including GPU scheduling and multi-node machine-learning workloads.
-
Practical experience with GPU clusters, cloud infrastructure and infrastructure-as-code such as Terraform or Pulumi.
-
Strong Linux and programming skills, particularly Python and Bash; Rust or C++ experience is an advantage.
-
Scientific computing or HPC experience is a strong positive, as is evidence of improving compute performance, cost or utilisation.