20% increase in sales performance thanks to a recommendation engine



format site sicara (Bannière moyenne (US) (Paysage)) (1)


We supported Beauvallet, a French SME specializing in meat production and distribution for restaurateurs, in their strategy to increase revenue. This was achieved by creating a recommendation engine that helps telesales agents boost daily order volumes by recommending similar products.

The Challenge

To successfully create this recommendation engine, it was necessary to add business rules to model the desires and needs of end customers. The construction of the end-to-end recommendation system involved data retrieval and cleaning, as well as displaying the algorithm's results on a web platform. The model is retrained automatically, and there is continuous iteration on the most relevant business rules.

The Solution

To facilitate the work of telesales agents, our teams developed an ALS recommendation engine. This involved creating a product matrix that allows the algorithm to score products, thereby suggesting items that customers might not have considered themselves.

Stack technique


The model was put into production, and all telesales agents use the algorithm daily.

10 000

Referenced items


Orders per day

Need expert advice?