Software Engineer (Compute Platform), London

Location: London, United Kingdom; hybrid work with attendance at the office 3 days a week

Company

hirify.global develops AI models and drug design systems to accelerate scientific discovery and the development of medicines.

What you will do

• Define the end-to-end GPU and TPU accelerator strategy, including infrastructure design, performance optimization, hardware integration, and cluster deployment.
• Build, monitor, manage, and support research, development, and production cloud infrastructure and systems.
• Drive infrastructure efficiency improvements up to the service layer used by machine learning platform teams.
• Improve the reliability and consistency of large-scale machine learning runs.
• Support hardware acquisition and deployment decisions and contribute to tooling, infrastructure, and architecture choices.
• Collaborate with machine learning, science, research, product, business development, and operations teams.

Requirements

• Real-world experience with large-scale AI and machine learning workloads.
• Experience designing cloud compute infrastructure, preferably on GCP.
• Strong programming skills.
• Significant experience deploying and working with Kubernetes.
• Familiarity with Nvidia GPU generations.
• Ability to work in a hybrid model and attend the London office 3 days per week.

Nice to have

• Background in machine learning software engineering or infrastructure SRE.
• Experience leading and delivering projects with multidisciplinary stakeholders.
• Familiarity with Google TPU generations, workload scheduling, hardware benchmarking, and machine learning efficiency research.
• Understanding of machine learning-driven research and development cycles.

Culture & Benefits

• Collaborative interdisciplinary environment spanning science, research, engineering, product, and operations.
• Values centered on thoughtful work, initiative, integrity, determination, and collaboration.
• Shared learning and an environment designed to support employees and diverse perspectives.
• Hybrid working model with regular in-person collaboration.