Foundation Model Engineer for Materials AI

CuspAI; Cambridge; Hybrid; GBP 90,000 - 130,000; Full time

Benefits offered by this job

Equity stake

28 days holiday (UK)

Parental leave

Professional development budget

Interdisciplinary teamwork

Job summary

CuspAI in Cambridge, London, Amsterdam or Berlin is hiring Research Engineers to train and evaluate foundation models for materials discovery. You own end-to-end workflows from data collection, through training at scale to integration, in a small team.

You will focus on representation learning over crystal structures and properties, LLM fine-tuning and post-training, or evaluation suites, collaborating with chemists and materials scientists to build scalable, sustainable AI for breakthrough

Qualifications

• Experience training or evaluating large models at scale.
• Strong software engineering skills and scalable systems.
• Hands-on experience with distributed training across GPUs.
• Interest in AI and materials science with ability to learn domain vocabulary.

Responsibilities

• Design and run scaling experiments for multi-modal foundation models.
• Lead parallel training of multi-GPU models and manage large data.
• Apply software engineering best practices for scalable, reliable systems.
• Own data curation and evaluation design for your models.

Skills

Model training

Distributed training

Representation learning

Computer vision

Materials science interest

PhD in CS/ML

PyTorch

Education

Master’s or PhD in CS/ML

Tools

PyTorch

Job description

CuspAI in Cambridge, London, Amsterdam or Berlin is hiring Research Engineers to train and evaluate foundation models for materials discovery. You own end-to-end workflows from data collection, through training at scale to integration, in a small team.

You will focus on representation learning over crystal structures and properties, LLM fine-tuning and post-training, or evaluation suites, collaborating with chemists and materials scientists to build scalable, sustainable AI for breakthrough