Utilizing Large Language Model Enabled Agents to Streamline Business Decision Making

The growing complexity and volume of communication in business environments significantly hinder timely decision-making processes. This paper addresses the problem of delays and inefficiencies in business decision-making by exploring how large language models (LLMs)-enabled agents to streamline communication and accelerate the business decision making process. Our project aims to develop and evaluate an LLM-enabled Agent as an innovative tool for semi-automating communication and decision making in business processes. The central research question asks how LLMs can be utilized to reduce communication complexities in business processes. Guided by a design science research framework, this project follows a structured artifact design, implementation, and evaluation process. The virtual environment simulates real-world conditions using synthesized business communication data like emails and meeting notes. The LLM-enabled Agent leverages Azure OpenAI services and integrates domain-specific customization to align the LLM’s outputs with business needs. Quantitative testing of the agent’s performance assesses its effectiveness in automating information gathering, document synthesis, and decision-making.