AI Agent Systems for Supply Chains: Structured Decision Prompts and Memory Retrieval

TL;DR

The paper tackles robust, adaptive inventory control in multi-echelon supply chains using LLM-based multi-agent systems. It first demonstrates that simple prompts with a fixed ordering and safety-stock strategy can yield optimal decisions in a restricted setting, then identifies limitations in cross-scenario performance. It then introduces AIM-RM, an adaptive agent architecture that retrieves and reuses similar past experiences via a memory module to coordinate decentralized agents more coherently. Empirical results show AIM-RM with RL logs delivering robust performance across diverse scenarios, approaching RL baselines while outperforming non-memory baselines, with insights on how prompt complexity and reasoning effort affect outcomes. These findings advance practical deployment of LLM-driven SCM agents and suggest directions for improving adaptability and handling demand dynamics in real-world settings.

Abstract

This study investigates large language model (LLM) -based multi-agent systems (MASs) as a promising approach to inventory management, which is a key component of supply chain management. Although these systems have gained considerable attention for their potential to address the challenges associated with typical inventory management methods, key uncertainties regarding their effectiveness persist. Specifically, it is unclear whether LLM-based MASs can consistently derive optimal ordering policies and adapt to diverse supply chain scenarios. To address these questions, we examine an LLM-based MAS with a fixed-ordering strategy prompt that encodes the stepwise processes of the problem setting and a safe-stock strategy commonly used in inventory management. Our empirical results demonstrate that, even without detailed prompt adjustments, an LLM-based MAS can determine optimal ordering decisions in a restricted scenario. To enhance adaptability, we propose a novel agent called AIM-RM, which leverages similar historical experiences through similarity matching. Our results show that AIM-RM outperforms benchmark methods across various supply chain scenarios, highlighting its robustness and adaptability.

Figures (9)

• Figure 1: Multi agent system for inventory management in a supply chain. At each tier, the agent places orders with the upstream agent and receives goods after a delivery lead time.
• Figure 2: Framework of the interaction between agents and the supply chain environment. Starting with the downstream agent, the agent at each tier observes its state and places an order with the upstream agent in a round. Our proposed agent retrieves historical experience data through similarity matching and then uses them when placing an order.
• Figure 3: Results of InvAgent (w/ step desc and safety stock strategy) with medium reasoning effort. These four graphs present, from left to right, the inventory, backlog, order quantities, and relative reward $(r / {\sf Opt}\xspace)$ within an episode. The “demand” legend corresponds to customer demand, which is plotted in the graph for order quantity.
• Figure 4: Results for the inventory level, backlog, order quantities, and cumulative reward within an episode. The graphs in the first line represent the results of AIM-RM (w/ RL log) with medium reasoning effort, and those in the second line represent the same model with a high reasoning effort.
• Figure 5: This prompt teaches the decision-making module input data including the current state and the output style.

Theorems & Definitions (3)

• Remark 1
• Remark 2
• Remark 3