Data Scientist II - Product in Reading

As our Data Scientist II - Product, you'll be a senior contributor and technical lead in our Data Science Product team. The Data Scientist II - Product takes ownership of the design, build and production operation of the data science products that underpin MarketCast's media and entertainment analytics. Operating at the intersection of quantitative market research and statistical modelling, this position leads solutions to moderately complex product problems, deploys and monitors models in production, and drives improvements to how the team works. Bayesian statistics is a core methodology for the team, and this role is expected to set the standard for Bayesian model design, validation and communication across products.

On a daily basis, the data scientist defines success criteria, manages projects with multiple stakeholders across the UK, US and India, and guides more junior team members in modelling and deployment. By delivering scalable, well-engineered products and raising the statistical capability of the team, this individual directly increases the value and reliability of the organisation's analytics.

We're Looking For:

• Solution Leadership: The individual takes the lead on directing, defining and implementing solutions to moderately complex product problems, effectively defining clear success criteria.
• Bayesian Modelling Leadership: The role designs, builds and validates production-grade Bayesian models, including hierarchical and multilevel models, Bayesian time series and forecasting, and probabilistic models of audience behaviour.
• Production Pipeline Ownership: The individual owns the design, development and maintenance of production modelling pipelines and products, ensuring they are robust, scalable and well documented.
• Deployment, Monitoring & MLOps: The role deploys Bayesian and machine learning models to production and evaluates their results over time, acting on model drift, data quality issues, calibration and performance degradation.
• Advanced Modelling & Innovation: The individual creatively addresses issues with models beyond basic feature engineering, drawing on in-depth knowledge across Bayesian, statistical and machine learning techniques.
• Continuous Improvement: The role critically evaluates where improvements are required in our products and ways of working, always through the perspective of business benefit to maximise value delivered.
• Stakeholder & Project Management: The individual independently manages internal stakeholders and expectations with minimal assistance, including Product, Market Research and Data Engineering teams.
• Mentorship & Team Enablement: The role guides more junior members of the team in Bayesian modelling, model development, code quality and deployment.
• Technical Communication & Reporting: The individual uses Python, SQL and Tableau to produce reporting and analysis, and presents technical work clearly to internal teams and clients.

Qualifications:

You'll typically have 5 or more years' experience in data science, with a proven track record of delivering data science products into production independently and of guiding more junior colleagues to do the same. We'd expect a bachelor's degree in Data Science, Statistics, Computer Science, Mathematics, or another highly quantitative field. A master's or PhD in a quantitative discipline is a plus, as is an equivalent practical academic foundation. Bayesian statistics will be a genuine area of depth for you.

You'll be fluent in hierarchical models, MCMC and approximate inference, prior elicitation, and model checking and comparison, and you'll have substantial hands-on experience in PyMC, Stan, NumPyro or similar. You'll know how to deploy, monitor and evaluate models in production, and you'll bring solid software engineering habits: version control, testing, code review and reproducible pipelines. You'll be genuinely proficient in Python, SQL and pandas. Experience with Polars is a strong plus.

You'll have proven experience with cloud platforms, ideally AWS, though Azure or GCP experience is also relevant. On the softer side, you'll be able to understand a business problem, draw conclusions from data and recommend actions, and you'll manage stakeholders and projects independently.

Who We Are:

MarketCast measures what moves people. Powered by AI, trained on decades of verified consumer response data, along with proprietary technology, MarketCast is the marketing effectiveness partner for the world's most ambitious brands.

Our Values:

• Move The Needle
• Client Success is Our Scoreboard
• Grow Every Day
• Own It

Benefits:

• Free movie tickets!
• 29 days annual leave PLUS Bank Holidays
• Flextime with core hours between 10am and 3pm
• 2 days’ work from home, per week
• 4% match pension scheme
• Enhanced maternity pay
• Regular social events in both UK locations
• Professional growth and career development including LinkedIn Learning