Lead Data Scientist

We’re looking for a hands-on, generalist data scientist to own ML across all parts of our broad business, including Membership, Food & Beverage, Operations, Digital and Events.

As Lead Data Scientist, you’ll be the go-to expert for turning our vast datasets into predictive and prescriptive products that change how we run our Houses and serve our members, from predicting member churn and personalising what members see, to forecasting demand so our Houses are staffed well.

You’ll take models from idea to production and keep them performing, always looking for new ways to create value for our members and the business.

You’ll also raise the bar on how the wider analytics team measures impact, so decisions across the business rest on sound evidence.

Machine learning products

• Own the design, build and continuous improvement of Soho House’s ML portfolio, including member churn and segmentations, personalisation and recommendation engines (e.g. events), and demand forecasting for labour and operational planning
• Frame business problems with stakeholders, pick the right approach (which may not always be ML), and set clear success metrics tied to commercial outcomes
• Own the full model lifecycle: scoping, development, validation, deployment, monitoring, scheduled recalibration and retirement

• Set up ML Ops standards: version control, model registry, documentation and governance
• Work with Data Engineering to put models into production on our stack (GCP, Snowflake, Airflow), with reliable scoring pipelines and outputs that downstream tools and teams can use
• Build monitoring and observability for model accuracy in production (tracked against agreed thresholds and baselines), data drift and business impact, with clear alerts and review cycles

• Set the standard for advanced measurement across the wider analytics team, educating analysts and acting as their go-to expert for things like experimentation and causal inference, and finding pragmatic, robust alternatives when the ideal method isn’t possible (e.g. where randomisation can’t be done)
• Create reusable frameworks, templates and guidance so analysts can run sound tests and impact measurement on their own
• Coach and upskill analysts, supporting the wider goal of building data literacy across the business

• Partner with Membership, Marketing/CRM, Digital, Operations and Finance to prioritise a roadmap of data science work by value
• Explain complex methods and uncertainty clearly to non-technical audiences, including senior leadership
• Direct and quality-assure work from offshore data science resources when workload requires: scope tasks, review code and models, and keep standards consistent

• Significant hands-on data science experience (typically 5+ years) across a broad range of problem types, such as classification and propensity, recommendation, time-series forecasting, clustering and segmentation
• A track record of putting models into production and owning them afterwards, not just building prototypes
• Strong Python (our preferred language) and its data science ecosystem
• A deep grounding in statistics, experimentation and causal inference (A/B testing, power analysis, variance reduction, difference-in-differences, synthetic control or similar)
• Excellent stakeholder skills: can turn a vague business question into a well-framed problem and explain results in plain English
• Comfortable as a senior hands-on IC who can also set direction for others and review their work without line-managing them

• Experience with a modern cloud data stack, ideally Snowflake (Snowpark, Cortex) and dbt
• Practical ML Ops experience, e.g. MLflow…