IMPLEMENTATION OF AN ML-ALGORITHM FOR PRODUCT RECOMMENDATIONS
GOALS AND OBJECTIVES
To increase the average purchase amount using targeted offers to customers.
To develop and implement a ML-system for recommendations at the time of checkout.
- A Machine Learning algorithms-based system.
Jet Infosystems specialists have developed and implemented an artificial intelligence-based system for the VEK ZHIVI pharmacy chain that analyzes customer purchase data with a machine learning model. Based on this analysis, the service shows the cashier three products, which the client is highly likely to add to their purchases, so that the cashier can provide a qualified recommendation. Products (together with their SKU number) are suggested from 27 thousand pharmaceutical products.
In order to cope with the task at hand, a whole set of Machine Learning methods was used, and the mathematical model was trained on a long-term data array storing customer receipts. So as to create a personal offer, the system analyzes several parameters simultaneously, including: the structure of the receipt, the list of goods purchased, and their prices.
Thus, based on comprehensive analysis of previous purchases, the ML-algorithm helps reveal buyers’ hidden needs to a high degree of accuracy, and suggests additional medicines at the time of any new purchase.
ML systems of this kind are now widely used in retail, both by retail chains and by online stores. To date, Jet Infosystems specialists have implemented more than 50 projects based on Machine Learning technologies for banks, retail, industry, insurance and other sectors.