Data Science at Tesco spans a variety of business areas and use cases, covering everything from online and stores through to logistics and trading.

In data science engineering, we aim to enable data scientists across the entire data science life cycle, from building new predictive models to getting them into a production setting and maintaining. We want to do this faster, better, cheaper. For this purpose we are building a Data Science Framework, where capabilities and components can be reused across projects. This framework spans across several platforms including an on prem big data Spark platform, cloud environments and edge. The primary language used by data science and engineering teams is Python.

In this talk, we will provide an overview of the platform, giving a detailed view of how this enables Tesco to deliver personalisation to our customers. This includes the development of a Feature Store, model training and evaluation and other aspects of our MLOps stack.