Automotive supply chains are pressured to improve efficiency through automation. During automotive assembly manufacturing, products are identified by a code and a brief, often ambiguous, textual label, which lacks semantic clarity to accurately identify product type. Expert operators must manually interpret bill of materials (BOMs) and 3D product models to deduce the component category along extensive supply chain, which is time-consuming and subjective. This paper presents an artificial intelligence (AI)-based conceptual framework to leverage large language models (LLMs) to automate the classification/clustering of finished goods/parts. By integrating ambiguous textual inputs with structured BOM data, this framework is the basis of LLMs for replicating expert reasoning for correct component class to improve categorization accuracy and processing time over traditional rule-based systems. This case study is applied to the automotive seat manufacturing supply chain.