Optimizing Service Operations via LLM-Powered Multi-Agent Simulation

Researchers introduce an LLM-powered multi-agent simulation framework for optimizing service operations by modeling human behavior through AI agents. The method uses prompts to embed design choices and extracts outcomes from LLM responses to create a controlled Markov chain model, showing superior performance in supply chain and contest design applications.

• →New LLM-powered multi-agent simulation framework addresses complex human behavior modeling in service system optimization.
• →The method treats optimization as stochastic with decision-dependent uncertainty using prompts to shape agent interactions.
• →On-trajectory learning algorithm simultaneously constructs gradient estimates and updates parameters in single simulation runs.
• →Framework outperformed traditional blackbox optimization and other LLM-based approaches in supply chain applications.
• →Case study with real behavioral data demonstrates both cost-effective evaluation and discovery of overlooked design solutions.