PROJECT & FUNDING SCHEMES

This project proposes the SLAssIstants— Multi-Agentic Generative AI-Driven Smart Logistics Assistants for B2B cross-border multi-stage multi-modal logistics. It aims to address three core pain points: (1) heavy customer service workload, (2) slow and inaccurate quotation planning, and (3) laborious decision model formulation under dynamic disruptions. The R&D methodology integrates LLM-based intelligent process automation, multi-agentic multi-modal analytics, and LLM-driven optimization model generation, supported by domain-specific fine-tuning, knowledge graphs, and Human-AI collaboration mechanisms. The platform will be validated through industrial pilots with Shenzhen S.C Global Logistics Co., Ltd. (hereinafter referred to as SCG) (the Customer AssIstant), THE TREE Global Co., Ltd. (hereinafter referred to as THE TREE) (the Quotation AssIstant), and Shenzhen Friend Supply Chain Management Co., Ltd. (hereinafter referred to as FRD) (the Decision AssIstant). Expected impacts include doubling customer service efficiency, reducing quotation lead time from 7 to 3 days, and cutting decision model development time by 60%. By embedding logistics expertise into generative AI, the SLAssIstants enable interpretable, adaptive, and real-time decision support—paving the way for scalable, human-centered AI adoption in global logistics.