> Markdown version of [/jobs/ext/3603242-data-scientist](https://www.wearedevelopers.com/jobs/ext/3603242-data-scientist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Data Scientist - **Company:** PureGym Limited - **Location:** London, UK - **Experience:** Expert - **Contract:** Permanent contract - **Skills:** Clean Code Principles, Python (Programming Language), Standard Sql, Management of Software Versions, Pyspark, Machine Learning Operations, Databricks - **Published:** October 7, 2026 - **Apply:** https://www.jobs.service.gov.uk/jobs/6ac3e5f7e7d4cf609bde8e4f/apply ## About the Role The most important thing is that you combine strong Data Science fundamentals with genuine geospatial expertise and can demonstrate how you've used modelling to influence real-world decisions., * A degree in a quantitative or geospatial discipline, or equivalent practical experience. * Several years of applied Data Science experience, including building geospatial models that have been used to inform real-world decisions. Experience within retail, leisure, property or real estate would be particularly relevant. * Strong geospatial modelling expertise, for example spatial statistics, gravity and Huff-style models, network and isochrone analysis and spatial autocorrelation. * Proven experience taking at least one model from prototype through to production, with an understanding of what it takes to maintain and monitor it once live. * Strong Python and SQL skills, with experience writing version-controlled, tested and maintainable code. * Experience with the Python geospatial ecosystem and working with PySpark on Databricks. * Working knowledge of dbt, MLflow and production model deployment patterns. * Confidence using AI-assisted development tooling appropriately as part of the development lifecycle. * Strong written and verbal communication skills, with the ability to explain complex technical concepts and model outputs clearly to non-technical audiences. ## Description Data Science at PureGym influences some of the biggest decisions we make, from where we open our next gyms and how we price, to how we keep millions of members coming back. Our models shape real decisions across retention, pricing, customer lifetime value and estate growth, with the impact showing up in the numbers our business runs on. We're a small, experienced team that's great at turning good ideas into working solutions quickly. You'll join people who care about doing Data Science well, and you'll see your work used to make real commercial decisions throughout the business, not just admired in a notebook. As our Senior Data Scientist, you'll take the technical lead on the modelling behind PureGym's spatial analytics capability. You'll help us answer some of our most commercially important questions: where should we grow, how do our gyms fit within their local communities, and what drives the performance of different locations? This is a role with significant ownership. You'll take spatial and location models from prototype through to trusted, production-ready assets, ensuring they're rigorous, explainable and capable of supporting confident investment and commercial decisions., * Own the spatial and location models that power our analytics tools, taking them from prototype through to maintained, well-documented production assets. * Build and evolve catchment, location and demand models that help us understand our markets, communities and opportunities for growth. * Partner closely with Property and Finance, translating commercial questions into analytical problems and ensuring our models deliver meaningful business value. * Rigorously validate models through out-of-time testing, leakage checks, sensitivity analysis and clear measures of uncertainty, ensuring every input genuinely reflects the information available at the point a decision is made. * Develop approaches for locations with limited historical data, using your technical judgement to determine when there is enough evidence to trust a result. * Work with Engineering and Data colleagues to turn model outputs into reusable, scalable data assets rather than one-off analysis or tables. * Partner with MLOps colleagues to deploy, monitor and retrain models, with clear ownership, versioning in MLflow and documented handover points. * Turn complex model outputs into clear, commercially relevant recommendations for senior stakeholders, including being transparent about uncertainty and where a model should, and should not, be used. * Evaluate external and third-party models with technical rigour, challenging assumptions and helping set the standard for how we assess them. * Contribute to shared development standards and peer review, sharing knowledge so that our capability can scale and no critical model is dependent on one person. * Help shape the future of spatial analytics at PureGym, identifying opportunities to improve our methods, tools and the commercial impact of our models. The Person You're an experienced Data Scientist with strong geospatial expertise who enjoys using complex data to solve real-world commercial problems where the answer isn't always obvious. You're curious enough to question the data and rigorous enough to challenge an impressive-looking result when the evidence doesn't support it. You're also pragmatic: you understand that the best model isn't necessarily the most sophisticated one, but the one that's robust, explainable and genuinely helps people make a better decision. You'll be: * Independent: You take ownership from problem definition and experimentation through to production, rather than stopping at proof of concept. * Commercially curious: You want to understand the decision behind the model and how your work ultimately creates value for the business. * Sceptical by default: You test assumptions, data and headline metrics before accepting them, whether they've come from an internal model or an external provider. * Pragmatic: You choose the simplest approach that produces a robust, reliable and explainable result rather than adding complexity for complexity's sake. * A strong collaborator: You enjoy solving problems alongside colleagues across Property, Finance, Engineering and Data. * A clear communicator: You're credible with senior non-technical stakeholders and can explain complex modelling, uncertainty and technical trade-offs in plain language.