Supply chain reconfiguration has emerged as a critical solution to disruptions. Knowledge Graphs (KG) provide an effective means of modeling and managing the data required for this reconfiguration. Moreover, maintaining human centricity in decision-making remains a key challenge in Supply Chain 5.0. This study proposes a Graph Retrieval-Augmented Generation (GraphRAG) approach using Large Language Models (LLMs) to facilitate human interaction with supply chain Knowledge Graphs. By enabling natural language queries, this method eliminates the need for technical query languages, allowing non-expert users to interact seamlessly with the KG. Our goal is to improve accessibility and promote human-driven decision-making in supply chain reconfiguration without requiring specialized computer programming skills. To validate this approach, we present a reconfigurable supply chain KG model coupled with an LLM for human interaction. Experimental results demonstrate the feasibility of this solution. Our methodology enhances human involvement by improving the overall interaction with the reconfigurable supply chain KG.