Remote ML Scientist: Multimodal Drug Discovery AI

Iambic; West of England; On-site; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Private medical insurance

Life assurance

Pension contributions

Flexible holiday allowances

UK Bristol office (optional)

Job summary

Iambic Therapeutics is seeking a Machine Learning Scientist to join the Enchant team. You will research and develop post-training methods for a multimodal transformer model trained on biomedical data, aiming to push the boundaries of foundation models in drug discovery.

You will design and evaluate post-training approaches, including supervised fine-tuning, RL, and parameter-efficient methods, while building scalable training infrastructure and collaborating across ML, software, and drug

Qualifications

• PhD or equivalent industry experience in a computational STEM field.
• Strong Python and PyTorch skills with end-to-end DL model work.
• Experience training large-scale transformer models and RL methods.

Responsibilities

• Research and develop post-training strategies for large multimodal models.
• Design RLHF, supervised/parameter-efficient fine-tuning objectives and evaluation protocols.
• Build scalable experimentation and optimization workflows for model configurations.
• Develop inference optimization and deployment support for discovery workflows.
• Collaborate with ML, software, and drug discovery teams to ground objectives in biology.
• Communicate results to internal teams and conferences; write high-quality code.

Skills

Python

PyTorch

Transformer models

Reinforcement learning

Experimentation

Education

PhD in machine learning, CS, computational chemistry, physics, or related field

Tools

Docker

Kubernetes

Weights & Biases

CUDA

Job description

Iambic Therapeutics is seeking a Machine Learning Scientist to join the Enchant team. You will research and develop post-training methods for a multimodal transformer model trained on biomedical data, aiming to push the boundaries of foundation models in drug discovery.

You will design and evaluate post-training approaches, including supervised fine-tuning, RL, and parameter-efficient methods, while building scalable training infrastructure and collaborating across ML, software, and drug