Data Scientist

Our forecasting models decide how much stock 1,200 retail sites hold. When they are wrong, it is expensive and very visible.

What you will do

• Own the demand forecasting pipeline end to end, from ingestion to the dashboard buyers actually use.
• Replace the three legacy models still running on rules nobody has revisited since 2021.
• Work directly with the buying team. They will tell you when your model is wrong before your metrics do.
• Publish an internal write-up for every model change that ships.

What we are looking for

• Strong Python and SQL, plus real experience with time series in production.
• The judgement to know when a simpler model is the right answer.
• The ability to explain a confidence interval to someone who does not want one.

Why on-site

The buying team sits in the Manchester office and the feedback loop is the job. This role is five days on-site and we are upfront about that.