Data Labelling Engineer (12-Month Fixed Term)

About CMR Surgical
At CMR Surgical, we're revolutionising the future of minimal access surgery with Versius, our next-generation surgical robotic system. We're growing fast, and we're on a mission to transform how surgery is delivered globally for the benefit of patients, clinicians, and healthcare systems.

Our values, Ambitious, Humble, People-Focused, and Fair & Responsible, define our approach to innovation, partnership, and growth. If you believe in making a meaningful difference in healthcare and want to be part of a purpose-driven company with bold ambitions, now is the time to join us.

CMR is an exciting and evolving company to be joining right now. Our commitment to our goal of building a world-class medical devices company is underpinned by the ground-breaking innovation of our next generation of surgical robots and a diverse team of talented and supportive individuals.

We're breaking new ground which makes working here exciting, it provides autonomy, opportunity and makes the things you do visible. Come and join us and see what you're capable of.

About the Role
We are looking for a Data Labelling Engineer to help build how we label surgical video in our Applied AI team. You'll create high-quality labelled datasets and build tools that make labelling faster, more consistent and easier to scale.

This is a hands-on role. You'll label and review surgical video yourself and use Python and existing AI models to automate as much of the work as possible. For example, models can suggest labels, carry them across video frames and point you to the footage that most needs a human check. You'll measure how much time this saves and whether quality holds up.

You'll work closely with senior engineers, clinical experts and our Quality and Regulatory team. You'll own well-defined pieces of work and keep clear records of how our datasets are built. Those records support model development and future regulatory submissions, such as to the FDA.

You don't need a clinical background. We'll help you learn the surgical terms you need, and qualified clinicians will make the final call on clinical questions.

This role is offered on a hybrid basis, with a minimum of 3 days a week being based in the office.

Responsibilities
Labelling and review

• Label and review surgical video, marking surgical phases and steps, instruments and anatomy. You'll use time-based labels, boxes and outlines, following agreed clinical definitions.
• Work with clinical experts to keep the labelling guidance and examples up to date, and help colleagues learn the tools.

Tooling and automation

• Write and maintain Python scripts and tool integrations to prepare video, sample frames, check metadata and move labels in and out of our systems.
• Use existing AI models to suggest labels and carry them across video frames. Check where they go wrong, for example when instruments are hidden, or the view is poor.
• Work with machine learning engineers to find the examples models struggle with and focus labelling effort there.

Quality and measurement

• Track how fast and how accurately labels are produced, including how often reviewers agree. Compare AI-assisted labelling with fully manual labelling before using it more widely.
• Summarise what each dataset covers and what's missing, and flag groups or conditions that aren't well represented.

Data governance and regulatory records

• Keep datasets versioned and traceable: where each video came from, how it was processed, what changed and which tool and model versions were used. Record training, review decisions and disagreements under approved procedures.
• Handle surgical video securely and follow our rules on confidentiality, anonymisation and access. Keep test data strictly separate from development data and keep AI suggestions separate from approved labels. Support blinded clinical reviews and prepare evidence for our Quality and Regulatory team.

About you
To be successful in this role, you will need:

• Practical Python skills for processing and automating data, using libraries such as NumPy, Pandas or OpenCV, plus Git for version control.
• Hands-on experience working with image or video data. That could be labelling, reviewing or curating datasets, and spotting and fixing quality issues. It can come from work, study or substantial personal projects.
• A working understanding of computer vision models such as detection, segmentation or tracking, including why training, tuning and test data must be kept separate.
• Careful, organised working habits. You document your work clearly, deliver defined tasks independently and ask for help when something is unfamiliar.
• Willingness to watch surgical footage regularly and to learn basic anatomy and surgical workflows, while handling sensitive information responsibly. No clinical background is needed.

Desired, but not essential:

• A degree in computer science, engineering, data science or a related STEM subject, or equivalent practical experience.
• Experience with annotation platforms such as CVAT or Label Studio, including their APIs, or comparable commercial or internal tools.
• Experience with AI-assisted annotation, active learning, video label propagation, foundation models or fine-tuning computer vision models.
• Familiarity with SQL, Linux, Docker, cloud storage or tools for dataset versioning and reproducible processing.
• Experience in medical imaging, robotics, autonomous driving or another domain where data quality and traceability are important.
• Exposure to controlled documentation, quality systems or dataset evidence supporting medical device submissions.

At CMR we're building a diverse, inclusive, and authentic workplace; what matters most to us is that you identify with our values and are passionate about making a difference.

Why Join Us?
We offer a competitive salary and a great benefits package including a bonus, pension, healthcare and enhanced global parental leave pay.

This role requires candidates to have eligibility to work within the UK. Strictly no agencies.