About Marquee

Marquee builds AI that helps professional football clubs make better decisions about recruitment, squad planning and match preparation. We combine football data, machine learning and generative AI to turn complex analysis into practical insights for coaches, analysts and recruitment teams.

The Role

We are looking for a Senior AI Scientist to develop the models behind our football intelligence platform. This is a hands-on individual contributor role with ownership of research projects from problem definition and experimentation through production and ongoing evaluation. You will work closely with our data, engineering, product and football experts on questions such as how a player may perform in a new league, how well they fit a team's playing style, and how teams can prepare for different opponents.

What you will do

• Lead applied AI research for player evaluation, performance forecasting, tactical fit and team analysis.
• Turn open football questions into testable hypotheses and clear modelling objectives.
• Design, train and improve models using match events, player and team histories, and spatial or tracking data where available.
• Build rigorous evaluations with credible baselines, tests on later seasons and unseen competitions, and explicit checks for data leakage, uncertainty and failure cases.
• Partner with engineers to bring models into production, monitor their performance and improve reliability, latency and inference cost.
• Collaborate on LLM and agent workflows that retrieve verified data and use model outputs to produce traceable football analysis.
• Help shape research priorities, review code and experiments, and share methods and findings with the team.

What you will bring

• 5+ years of relevant experience in machine learning, AI research or data science, or equivalent depth through substantial applied research, with evidence of independently delivering complex projects.
• Strong Python skills and practical experience with PyTorch, SQL and scientific computing tools such as pandas and NumPy.
• Solid foundations in machine learning, deep learning, statistics and experimental design, with the judgement to select methods that fit the problem.
• Experience building and evaluating predictive models using large, imperfect datasets, including an understanding of missing data, selection bias and changing data distributions.
• Strong background in data structures, algorithms, and machine learning fundamentals, with a deep understanding of data science methodologies and best practices.
• Experience working with complex, high-dimensional football and sports datasets, including event data, tracking data, and physical/biometric metrics.
• Proficiency in building automated data pipelines, ETL processes, and validation checks for heterogeneous external and internal data sources.
• A track record of turning research into working products, with reproducible experiments, maintainable code and collaboration with engineering teams.
• The ability to explain technical findings and limitations clearly, work independently and learn from product colleagues and domain experts.

Nice to have

• An MSc or PhD in computer science, mathematics, statistics, engineering or a related field.
• Experience with transformers, sequence modelling, representation learning or self-supervised learning.
• Experience building and evaluating LLM applications, retrieval systems or tool-using agents.
• Experience with ranking and recommendation systems, probabilistic modelling or causal inference.
• Experience with sports analytics, event or tracking data, or a strong interest in football.

Why Marquee

You will own research that reaches professional football clubs and work closely with the people who use it. The role offers room to explore new methods, influence the product and judge success by whether your models help clubs make better decisions.

How to apply

Show us research that reached real use. Please include your CV and a short example of an AI project you took from research to practical use, explaining the problem, your contribution and how you evaluated the result.