Sonia Foco
Design and Evaluation of a Text-to-SQL Agentic System for On-Shelf Availability Analytics.
Rel. Paolo Garza. Politecnico di Torino, Corso di laurea magistrale in Data Science And Engineering, 2026
Abstract
Supply chain processes heavily rely on the ability to extract timely and actionable insights from large volumes of operational data. In the Fast Moving Consumer Goods (FMCG) sector, the monitoring of On-Shelf Availability (OSA) represents a critical analytical task, as product unavailability directly affects sales performance, customer satisfaction, and brand loyalty. However, traditional Business Intelligence (BI) dashboards are often data-dense and require considerable time for navigation and interpretation, making it challenging to quickly identify the most critical gaps and corrective actions. Recent advances in Large Language Models (LLMs) have opened new opportunities for retrieving data using natural language, particularly through Text-to-SQL systems that translate user questions into executable database queries.
This thesis presents the design, implementation, and evaluation of an agentic Text-to-SQL system that enables supply chain analysts to interact with an enterprise data warehouse through natural language, returning both scoped SQL queries and business-oriented narrative summaries
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Informazioni aggiuntive
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
