<![CDATA[Generative Artificial Intelligence (GenAI), particularly large language models (LLMs), is increasingly explored to strengthen decision support in supply chain and inventory management by improving interpretability and access to analytics. However, prior work is scattered across optimization, simulation, logistics, and governance discussions, limiting clear system design guidance. This study conducts a Systematic Literature Review (SLR) following PRISMA 2020 across IEEE Xplore, ACM Digital Library, ScienceDirect, SpringerLink, and Google Scholar, yielding 200 records, of which 34 studies were included in qualitative synthesis. Results show that LLMs are predominantly positioned as orchestration and explanatory layers operating alongside structured components such as optimization solvers, simulation engines, and digital twins, rather than as autonomous decision-makers. Governance, organizational readiness, and trust emerge as central considerations for operational deployment. This review provides an evidence map linking LLM roles and integration architectures across supply chain and inventory contexts. While LLMs offer strong augmentation capabilities, direct empirical validation for specific contexts such as web-based inventory systems remains limited; design implications for such systems are derived from the broader corpus, underscoring the need for standardized evaluation benchmarks and targeted empirical studies.]]>
Copyrights © 2026
