Empowering industry with Graph RAG: leveraging Generative AI for data insights and decision making

this research examines the role of Graph Retrieval-Augmented Generation (Graph RAG) within production management, utilizing a dataset focused on the manufacturing processes in the furniture industry. The dataset collects data on furniture production, including information on production orders, involved departments, processing times, quantities produced and associated sales orders. The growing complexity of production processes makes the adoption of advanced tools, including Artificial Intelligence (AI), increasingly crucial to support management in making informed and timely decisions. In this context, Graph RAG represents a significant innovation, as it allows users to query large datasets through natural language queries, simplifying access to key information and reducing the time required to obtain strategic insights. The analysis focuses on two main aspects: first, it explores how the adoption of Graph RAG can support both operational and strategic decision-making by streamlining the retrieval and interpretation of production data; second, it evaluates the reliability of this approach compared to traditional spreadsheet based analysis, with a focus on accuracy and time efficiency. The evaluation involved a set of 50 questions of varying complexity, used to compare the performance of a human analyst with that of the Graph RAG system. This study offers a preliminary contribution to understanding how emerging AI-based retrieval methods can enhance responsiveness and analytical capability in manufacturing environments.