ML Engineer

Who we are

Our mission is to transform how healthcare organisations work together with their workforce. Our Connected Scheduling™ platform connects healthcare organisations and their staff giving them more autonomy and control on how and when they work. Over 50% of UK GP practices use Lantum, and over 30% of UK hospitals rely on Lantum workforce products. We have developed a completely new approach to scheduling staff using AI to balance the vast amounts of complexities in workforce scheduling and we have seen game‑changing results.

We have not only saved millions for the NHS, but we have countless stories of how we have improved the lives of clinicians who, for the first time, are able to plan their work lives around their personal lives.

What sets us apart is not only our leading edge technology and approach to innovation, it’s our culture and our strength of mission. Our incredible team is the driving force behind our success and this propels our competitive edge. We are diverse (10+ nationalities and 53% female workforce), we are authentic and true to ourselves, we are creative and focused and we work hard together to change our industry.

Our team is supported to deliver their best work with clear career progression and a strong feedback culture.

We have a bright and modern office which you can work from throughout the week and 3 core office days per week (Monday, Tuesday & Wednesday) where the whole team comes together.

About the role

This role strengthens the core of our AI scheduling engine. You will build and optimise the models that power Connected Scheduling, improve our internal data science capability, and work closely with engineering to deliver fast, accurate and reliable solving at scale.

Responsibilities

• Build, optimise, and maintain production‑grade AI models for complex rota scheduling
• Improve data pipelines, workflows, and experimentation processes to enhance model reliability
• Collaborate with engineering to embed AI into core product workflows
• Apply scientific best practice to ensure accuracy, fairness and compliance across all models

About You - We’ll be looking for

General

Our ideal candidate is an individual who has:

• Strong end‑to‑end data science skills with experience deploying models into production
• Deep expertise in Python, ML frameworks, optimisation methods and cloud engineering
• A scientific, hypothesis‑driven mindset with high attention to accuracy and rigour
• Ability to work with messy real‑world data and design robust solutions
• Clear communicator who can work effectively with engineering and product teams

Education and Training

Our ideal candidate is an individual who has:

• A degree (Masters and/or PhD preferred but not required) in a numerical field such as mathematics, statistics, physics, computer science, engineering or another STEM‑oriented subject
• Demonstrable experience in delivering production‑grade code
• Some formal training in (or comparable deep practical exposure to) descriptive statistics, probability, inferential statistics, software development and general data science fundamentals.

Technical Experience

An ideal candidate has demonstrable skills and experience in the following technologies.

Required (ideally most of the following):

• The wider Python (3) data science stack and ecosystem (such as Pandas, Num Py, Jupyter notebooks, Sci Py, FastAPI, Flask, Matplotlib and similar)
• Core ML and DL frameworks (such as PyTorch (strongly preferred), Keras, Tensor Flow, scikit‑learn and similar)
• Cloud compute, infrastructure, services and deployment wrt. end‑to‑end data science (ideally AWS (such as S3, EC2, Lambda, ECR, ECS))
• Data visualisation methods and tools (such as Matplotlib, Bokeh or Seaborn)
• CI/CD
• Git
• An appreciation for solid coding practices

Prior exposure to or interest in some of the following is highly beneficial:

• Constraint/constrained optimisation and programming (particularly using meta heuristics for scheduling problems) in relation to both practical solvers and formal theory
• Opta Planner/Time Fold or Google OR‑Tools
• Basic containerisation via Docker
• MLOps platforms, services and tools (such as DVC, MLflow, Sage Maker or Weights & Biases)
• Agentic applications and/or…