WaveAccess and Cloud.ru have launched a joint AI solution designed to forecast demand for individual menu items in restaurant chains. The system aims to optimize inventory planning, reduce food waste, and minimize stockouts during peak hours.
During a pilot project for a fast-food franchise, the AI model achieved a 97% forecast accuracy. The implementation resulted in a 14% reduction in food losses compared to manual planning methods.
The model utilizes historical sales data and incorporates variables such as seasonality, weather, promotions, and local events. It generates hourly forecasts for specific locations, running on Cloud.ru GPU infrastructure.
WaveAccess manages the end-to-end implementation, including integration with platforms like R-Keeper, iiko, and 1C. This allows clients to deploy AI without maintaining internal data science teams.
The partnership designates WaveAccess as a technology partner within the Cloud.ru ecosystem. Cloud.ru provides scalable GPU resources, while WaveAccess handles consulting, model development, and system integration.