Answers are grounded in this paper and cite their sources.
As Industry 4.0 is coming, Manufacturers need to evaluate suppliers with limited staff and incomplete information. Smaller suppliers may be particularly difficult to assess: their records are less standardized, and limited access to finance can obscure their capacity to fulfil orders. This paper examines whether Financial Large Language Models (FinLLMs), including FinGPT and BloombergGPT, can assist Supplier Relationship Management (SRM). It synthesizes research on quality-management workloads, supplier credit assessment, document processing, and supplier discovery, and considers the applications and limitations of the two models as representative cases. The review identifies four possible uses: extracting risk signals from text, combining textual and relational data for credit assessment, reading engineering documents, and finding suppliers through graph-based search. It also identifies limits in process-specific knowledge, numerical reasoning, confidence calibration, and the currency and coverage of training data. To clarify who should act on model outputs, the paper proposes a "boss-employee" framework organized around task division, decision authority, and calibration through human review. The framework treats AI as support for procurement staff rather than a substitute for their final judgement. It remains conceptual and requires testing in manufacturing settings.
No takes yet. Share an insight, caveat, or question.
Zunhe Xu (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: