Machine Learning Engineer — Scientific ML / Biopharma
Remote: EU/UK based
We are supporting a search for a Machine Learning Engineer to join a scientific software business developing advanced modelling technology for biopharmaceutical process development and manufacturing.
This is an applied ML engineering role sitting at an interesting point between scientific research and production software.
The organisation has Process Model Scientists developing and validating sophisticated modelling approaches. The challenge is then taking that work and turning it into reliable, maintainable, reusable production ML code that can become part of a scalable software product.
That is where this person will sit.
The role is deliberately not positioned as pure model research, generic backend engineering, or MLOps platform ownership. Instead, it is focused on productionising validated scientific and modelling work while maintaining the scientific integrity of the underlying models.
This is important because model quality within this environment goes beyond accuracy or loss. Changes to features, transformations, dependencies, modelling logic, or production code can alter biological constraints and scientifically important outputs even when conventional ML metrics appear healthy.
The successful person therefore needs the software-engineering discipline to build production systems alongside enough mathematical and ML depth to understand what the models are doing and recognise when changes introduce scientific regression risk.
The role will focus on:
- Productionising validated modelling work from Process Model Scientists
- Integrating ML and hybrid models into production workflows
- Improving the structure and maintainability of the ML codebase
- Designing reusable abstractions for recurring modelling patterns
- Implementing feature transformations alongside scientific teams
- Building internal evaluation, reporting, and validation tooling
- Introducing new ML libraries, modules, and engineering approaches where useful
- Investigating regression when model behaviour or scientific outputs change
- Debugging model-related workflow issues
- Working across orchestration, artifact tracking, and experimentation where ML code intersects with the wider platform
- Reducing the software-engineering burden on Process Model Scientists
We are looking for someone with:
- Strong Python and software-engineering skills
- Experience building production ML, AI, or model-driven applications
- A strong understanding of training, evaluation, inference, validation, and regression risk
- Experience with PyTorch, TensorFlow, Scikit-learn, JAX, or comparable frameworks
- The ability to reason about features, model behaviour, validation logic, metrics, and scientific constraints
- Familiarity with technologies such as Flyte, Kubeflow, MLflow, Ray, Optuna, or similar
- Good systems-design judgment
- Strong STEM foundations and comfort operating in a mathematically complex environment
- High autonomy and an ability to challenge, improve, and standardise existing approaches
- Ideally 3+ years of relevant experience, although the company is open to fewer years where the underlying foundations are strong
Previous biotech experience is not essential. Experience with scientific ML, hybrid modelling, simulation, industrial ML, or working directly with scientists would be particularly relevant.
This is an excellent opportunity for someone who wants to build a career around serious applied ML rather than purely consumer-facing AI applications. You will work alongside scientists and engineers on technically demanding problems and help create the engineering layer that turns advanced modelling ideas into scalable products.
Apply here or if you know someone from your network, please let me know.
