VMES Chatbot: Leveraging Large Language Models for SQL-Based Information Retrieval

AI-driven chatbots have become a powerful tool in educational settings due to the increasing demand to optimize academic information retrieval. This paper introduces Vincent Mary School of Engineering, Science and Technology chatbot (VMES chatbot), a conversational AI system that leverages large language models (LLMs) and Text-to-SQL generation to automate course-related queries. In order to ensure rapid and reliable database interactions, the chatbot uses few-shot prompting to convert user questions into optimized SQL queries. Furthermore, a hybrid agent-based approach improves adaptability across different academic information systems by enabling flexible querying of relational databases and structured data. The system offers faculty members and students an easy-to-use platform by integrating with an API and web interface. The potential of LLM-powered agents is demonstrated by VMES Chatbot, which enhances accessibility to academic resources by combining natural language understanding, structured data processing, and query optimization in university information retrieval systems.