Order for demand.
Not just a gut feeling.
A cloud-based forecasting system learns from sales data to help restaurant managers order fresh products with a better view of demand.
A restaurant chain that delivers fresh products to its restaurants
The challenge
A restaurant chain delivers fresh products to its restaurants. One of its challenges is to understand food demand in its stores and restaurants.
Twice a week, restaurant managers have to order products for next-day delivery. Anticipating what their customers will need is a tricky task for any manager.
The goal: help restaurant managers forecast demand, and consistently strike a good balance between food waste and shortages.
The solution
To better predict the stock needed in stores and restaurants, we used machine learning to learn from sales ticket data.
In collaboration with IDLabs, we developed a deep learning neural network that predicts the quantity of each product an individual restaurant should order for the coming days.
The whole solution was fully integrated and automated in a cloud environment (Microsoft Azure).
The result
Thanks to the fully automated solution, food demand in the restaurants can be forecast more accurately.
The result is more than a tool to assist restaurant managers: the solution allows for less food waste and more revenue.