Machine Learning Engineer

PureGym Limited; City of Westminster; Hybrid; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Free nationwide gym membership for you

Hybrid working

Flexible working culture

Personal private healthcare including

Life insurance x4

Company pension contribution

25 days annual leave plus 1 personal

Option to purchase additional holiday

Learning & development resources

Enhanced maternity/paternity/adoption

Job summary

PureGym Limited is seeking an engineer to take machine learning models from notebook to reliable, scalable production services, owning deployment, monitoring and retraining across the model estate. You will build reusable components, tooling and standards to ship models faster and more cost-effectively, working closely with Data Scientists and Data Engineers to shape how we do MLOps at PureGym.

You will implement robust CI/CD pipelines, monitor data drift and model performance, and automate

Qualifications

• Strong Python and SQL skills with software engineering practices.
• Hands-on experience taking ML models into production and supporting them live.
• Experience with MLflow or similar model lifecycle tooling.
• Experience building or working with CI/CD pipelines (GitHub Actions / Azure DevOps).
• Clear communication with technical and non-technical colleagues.
• Degree in Computing, Engineering or another quantitative discipline or equivalent experience.

Responsibilities

• Own the path to production for ML models including packaging, deployment, release management and rollback.
• Build robust CI/CD pipelines for models and data science code with automated testing and gates.
• Monitor data drift, model performance and pipeline health with alerting and runbooks.
• Automate recurring model runs, retraining and scoring to free Data Scientists for new work.
• Create reusable production components from model definitions on the semantic layer.
• Help define and evolve MLOps standards including versioning and documentation.
• Collaborate with Data Engineers on upstream pipelines and with Data Scientists on model design.
• Participate in code reviews and coach Data Scientists in software engineering best practices.
• Monitor cost and performance of the model estate and improve efficiency.
• Shape how ML engineering is done at PureGym and develop scalable patterns.

Skills

Python

SQL

Software engineering practices

CI/CD

Testing

Code review

Version control

AI tooling

Communication

Practical ML production experience

Education

Degree in Computing/Engineering/quantitative field

Tools

MLflow

GitHub Actions

Azure DevOps

Databricks

Unity Catalog (adv)

Job description

You'll be the engineer who takes our machine learning models from notebook to reliable, scalable production services, and keeps them performing once they're live. This is a hands-on role with genuine ownership. You'll own deployment, monitoring and retraining across our model estate, while building the reusable components, tooling and standards that make every new model faster, safer and more cost-effective to ship than the last. Working closely with our Data Scientists and Data Engineers, you'll help shape how we do MLOps at PureGym, bringing strong software engineering practices into the modelling lifecycle and making it easier for the team to get great models into production.

What you’ll be doing:

• Own the path to production for our machine learning models, including packaging, deployment, release management and rollback using Databricks and MLflow.
• Build robust CI/CD pipelines for models and data science code, introducing automated testing, environments and appropriate release gates.
• Build monitoring for data drift, model performance and pipeline health, with clear alerting and runbooks so issues can be identified and resolved quickly.
• Automate recurring model runs, retraining and scoring that are currently manual, freeing our Data Scientists to focus on new modelling and experimentation.
• Turn models and metric definitions created by Data Scientists into robust, reusable production components on our shared semantic layer.
• Help define and evolve our MLOps standards, covering versioning, model registry practices, documentation and handover.
• Partner with Data Engineers on upstream pipelines and Data Scientists on model design, ensuring production requirements are considered from the outset.
• Take an active role in code reviews and coach Data Scientists in software engineering best practice.
• Monitor the cost and performance of our model estate and identify opportunities to make it more efficient.
• Help shape how machine learning engineering is done at PureGym, creating patterns and standards that can scale as our use of ML grows.

The Person You're an engineer who enjoys the challenge of making machine learning work reliably in the real world. You don’t just want to deploy a model and move on. You care about what happens afterwards: how it performs, how it’s monitored, how easily it can be changed and how confidently the next model can be taken into production.

The most important thing is that you've successfully taken machine learning models into production and understand what it takes to keep them running reliably.

Required skills and experience

• Strong Python and SQL skills, combined with good software engineering practices including testing, packaging, code review and version control.
• Experience with AI assisted tooling
• Hands‑on experience taking machine learning models into production and supporting them once live.
• Experience with MLflow or similar model lifecycle tooling, including deploying models for batch and/or near real-time use cases.
• Experience building or working with CI/CD pipelines, for example GitHub Actions or Azure DevOps.
• The ability to communicate clearly with both technical and non-technical colleagues.
• A degree in Computing, Engineering, a quantitative discipline or similar or equivalent practical experience

Advantageous skills you may bring

• Experience working with Databricks, ideally including Unity Catalog, Jobs/Workflows and PySpark.
• An understanding of infrastructure as code and reproducible deployment practices.
• Familiarity with dbt and modern data warehouse/lakehouse patterns.

The PureGym Group is a global gym business with a community of more than 700 gyms and over 2 million members. At PureGym, everything we do comes back to one simple idea: helping you feel good.

We’re proud of our people and have a strong focus on internal progression. Championing diversity, we are committed to providing an excellent employee experience and workplace culture. As such, we are a Disability Confident Committed employer, and offer interviews to candidates who meet the essential criteria for a role, and opt‑in to the scheme on their application form. PureGym is proud to be an equal opportunities employer. Our company mantra is 'Everybody Welcome' and we are dedicated to promoting a diverse and inclusive place to work. From a hiring standpoint, we welcome applicants from all backgrounds and are committed to ensuring that our PureGym colleagues reflect the diversity of the nation as well as our millions of gym members we serve.

Benefits

• Free nationwide gym membership for you + 1
• Hybrid working
• A truly flexible working culture
• Personal private healthcare, including digital GP
• Life insurance x4
• Company pension contribution
• 25 days annual leave, plus 1 personal day
• Option to purchase additional holiday (up to 5 days)
• Great learning & development resources
• Enhanced maternity pay, paternity and adoption leave, Independent: Comfortable taking ownership and setting the engineering pattern rather than waiting for one to be defined.

Our core values / qualities we value

• Pragmatic: You favour the simplest robust solution over unnecessary complexity.
• Collaborative: You enjoy working across Data Science, Data Engineering and Analytics and can explain technical trade‑offs clearly.
• Quality‑focused: You care about reliability, maintainability and leaving good documentation behind for the next person.
• Curious: You want to understand the wider PureGym business and how Data Science creates value, not just the modelling lifecycle.