Demand Forecasting Engineer — Fashion Retail

Isabella Gomez

Freelance · Mid-level · London

This role is no longer accepting applications

Similar open roles you can apply to now:

About the role

We sell clothes. Fast ones. Our buying cycle is six weeks from trend identification to floor. We carry 4,800 SKUs in season across womenswear, menswear, and accessories, and we turn our inventory four times a year. Our current demand forecasting model is a spreadsheet. It was built by someone who has since left. Our head of buying has been maintaining it by adding rows. It does not account for trend velocity, it cannot distinguish between a product that sold out because it was popular and one that sold out because we underordered it, and it has no mechanism for incorporating real-time sell-through data from our Shopify store and three wholesale accounts. We are commissioning a ten-week project to replace it with something better. We are not asking for a machine learning system that requires a data engineer to operate. We are asking for a forecasting model that is accurate, understandable to a buyer with no data background, and maintainable without specialised engineering support. If you can build that and explain it clearly, the buyer who uses it every day will love you for it. There is also a strong possibility of ongoing work maintaining and improving the model through seasonal buying cycles.

Key Deliverables

List the expected deliverables for this project.

Audit the existing spreadsheet-based forecast and identify the highest-cost failure modes in terms of over-stock and under-stock decisions
Build a replacement demand forecasting model incorporating trend velocity signals, real-time sell-through, and promotional calendar inputs
Develop a Tableau dashboard that presents current season forecasts and sell-through performance in a format usable by the buying team
Document the model logic in plain language — the model must be understandable and explainable by someone without a data background
Conduct a structured handover session with the head of buying at the end of the engagement

Requirements

Technical stack needed for this mission.

3+ years of data science or analytics with specific experience in demand forecasting or inventory planning
Python for data processing, model development, and building a lightweight pipeline that runs without engineering oversight
Scikit-learn for time-series forecasting, regression, and model evaluation — classic statistical methods often outperform complex ones for retail demand, and we want someone who knows why
SQL for extracting sell-through, inventory position, and returns data from our Shopify and wholesale data exports
Statistics: seasonality decomposition, trend modelling, and the ability to explain forecast intervals in terms a buyer can use to make purchasing decisions
Tableau for producing sell-through dashboards and forecast visualisations that the commercial team can navigate independently