Research Software Engineer – Foundation Models for Science

Key Skills

AI technology, Cloud computing, Deep Learning, English Proficiency, Foundation Models, High performance computing- HPC, Kubernetes-K8s, Prototypes

Job Description

IBM Research, in the UK, is seeking high calibre candidates with a proven track record in applied research and development, to work on research that makes a difference. We offer a stimulating research environment comprising multi-disciplinary teams with diverse expertise working on challenging projects in a world-class research organization. We have access to cutting-edge experimental environments and strong ties with industrial, national, and academic research labs around the globe. You will want to be part of a team that is making a difference on how we can accelerate discovery using advanced technology.

Your Role and Responsibilities

The work location for this role is either Daresbury Labs or Hursley Labs for a minimum of 3 days a week (inclusive of Tuesdays and Thursdays).

Foundation models are transforming what we think of as possible in a wide range of areas. This role will be part of a team investigating the use of foundational models (FMs) into science. Key questions that you might be looking to answer are:

• How do we deploy FMs in such a way to maximise impact for scientific users?
• How do we optimize the utilisation of FMs with, for example, simulation workflows
• How do we integrate FM training and inference into wider-scope computational scientific workflows

Key outputs of this role are readily reusable software stacks, exciting real-world demonstrators and feedback into our existing FM research areas.

Required Technical and Professional Expertise

• A knowledge of key concepts in the FM literature, and expertise in building and deploying them for real-world examples.
• A knowledge of cloud technologies, specifically Kubernetes and expertise in leveraging them for large-scale AI or FM workloads.
• Ability to identify fundamental problems from real-world cloud use-cases and to design, build and validate successful cloud-focused solutions.
• Ability to demonstrate and evaluate systems/software via experimental method, particularly through hands-on creation of prototypes.
• For example configuring large-scale server/networking/storage infrastructures; deploying research prototypes within production-grade HPC/Cloud infrastructures; deep contribution or ownership of an opensource cloud middleware framework.
• Strong communication skills and the ability to collaborate effectively within local team
• Excellent command of the English language, both verbal and written

Preferred Technical and Professional Expertise

• Expertise and experience in deep learning applied to physical science problems, with a particular focus on chemistry and materials science.
• Strong contribution record with either publications in top peer-reviewed scientific conferences and journals or strong leadership track-record in opens source communities, with a particular focus on foundation models for science, or large scale machine learning models for science.
• Track record in being part of highly collaborative teams to tackle important problems which produce high impact solutions.