ML Engineer

Most diseases are still poorly understood at a biological level. Despite decades of research, the causal mechanisms driving many conditions remain unclear, limiting our ability to identify the right targets, design the right interventions and bring the right medicines to patients.

The AI Accelerator exists to change that. Based in London and sitting within Computational Innovation (@computationalinnovation), a global organisation spanning computational biology, human genetics, data excellence and AI, the Accelerator’s mission is to build production-quality AI capabilities that deepen our understanding of disease biology and increase probability of success.

We do this by applying neural-based methods across the biomedical data landscape to integrate heterogeneous, multimodal data sources, infer biological relationships and embed causal thinking into what we build. The goal is not just to predict but to explain and understand why disease occurs.

It could be electronic health records and medical imaging to support patient segmentation. It could be ‘omics data to identify novel therapeutic targets. It could be predicting transcriptional change for a given disease-causing variant. It could be simulating the effect of modulating a target of interest.

A core component of the AI Accelerator is AI Systems, a team focused on designing, building and deploying multimodal foundation models across the vast biomedical data landscape that will be used within Computational Innovation to enhance and accelerate portfolio decision-making.

We are looking for an ML Engineer to join the AI Systems team and contribute to work at the frontier of biomedical AI. This is a hands-on engineering role with real stakes, as the models you help build will be used to make decisions about which indications to pursue, in which patient population and against which target.

You will work in close partnership with AI scientists and more experienced ML engineers, supporting the translation of validated research prototypes and architectural designs into production-ready implementations at a high engineering standard. You will contribute to early architectural discussions, offering engineering perspectives on training, efficiency, scalability and production-readiness, and iterating with the team on design decisions under senior guidance.

This is a role for someone who takes pride in engineering craft, who writes clean, well-tested, well-documented code, and who cares that biology retains its integrity as models move from research to production. Your engineering work will go far beyond the model card; it will connect directly to human health outcomes.

Contribute engineering perspectives to architectural design discussions, supporting decisions that ensure foundation models train and infer efficiently and are scalable

Under the guidance of senior ML engineers, implement biomedical foundation model components such as training code, data loaders, tokenisers, inference logic and fine-tuning interfaces to a high engineering standard

Work closely with AI scientists to help translate validated research prototypes into robust, production-quality model artefacts, and contribute to benchmarking and performance evaluation

Contribute to the optimisation of validated models and inference pipelines, applying techniques such as quantisation, distillation and pruning to help meet production efficiency and latency requirements without compromising model performance

Write clean, well-tested, well-documented code and follow the engineering standards set by the team

Support model handovers to MLOps engineers, contributing documentation covering capabilities, known limitations, failure modes and retraining criteria

Stay current with advances in ML engineering, distributed training and biomedical AI tooling

Postgraduate degree in Machine Learning, Computer Science, Computational Biology or a related technical field; PhD preferred or MSc with equivalent industry experience

Hands-on experience with deep learning and foundation model implementations such as transformers, pre-training and fine-tuning

Some experience contributing to production-quality model artefacts, with a growing understanding of what is required to move from research prototypes to reliable deployment

Experience collaborating with researchers during the implementation process

Proficiency in Python and deep learning frameworks such as PyTorch

Strong software engineering fundamentals such as writing clean, testable, well-documented and maintainable code, version control, code reviews

Working knowledge of distributed training frameworks such as PyTorch Distributed, DeepSpeed, FSDP or Ray Train

Exposure to model optimisation techniques for inference, e.g. quantisation, distillation, pruning

Experience working with biomedical data modalities such as genomics, multi-omics, clinical or imaging data in an ML context is advantageous

Publications or contributions to open-source ML projects or tooling

This is a hybrid role with approximately 3 days a week in the office

Boehringer Ingelheim has been recognised as a Top Employer in the UK, demonstrating our commitment to building an exceptional workplace through strong people practices and supportive HR policies.

To learn more about why BI is a great place to work, visit:

https://www.boehringer-ingelheim.co.uk/careers/uk-careers/why-great-place-work