The project, developed under the Seal of Excellence DIPS (Digitalization and Innovation of Public Services) initiative and carried out in collaboration with the NLP Research Unit, supports Considi S.p.A. in validating and adopting advanced AI solutions for supply chain management. It focuses on assessing the reliability and impact of AI-based systems for forecasting scenario generation, enabling more informed, data-driven decision-making.
A structured experimentation framework validates a chat-based solution integrating machine learning for time series forecasting with Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG). This enables contextual analysis of complex factors influencing sales channels and supports the generation of alternative forecasting scenarios. By combining predictive analytics with language-based models, the project aims to enhance the adaptability, resilience, and efficiency of industrial supply chains.
This service is implemented within the DIPS ecosystem as part of the “Test before Invest” initiative, which supports organizations in safely testing Artificial Intelligence technologies and methodologies before committing to significant investments.
