Developing a User-Friendly Conversational AI Assistant for University Using Ollama and LLama3
Jayrajsinh Gohil, Hagos L. Shifare, Madhu Shukla · 2025
This paper presents the development and implementation of an AI-driven chatbot for a university, designed to assist students and parents by providing accurate and instant information about the university. The chatbot leverages advanced LLM technologies, specifically LLama3, integrated with Ollama, to ensure high-quality natural language processing. Using LangChain, the system processes user queries by combining them with a structured prompt template, generating precise responses, and storing interactions in a database for future reference. The architecture includes modules like login, signup, and chat, ensuring a user-friendly interface. The chatbot's capability to reliably give precise responses was evaluated both manually and with volunteers in actual situations. This project showcases the uniqueness of utilizing Ollama and LLama3, which distinguishes it from conventional solutions by providing incredibly trustworthy and relevant replies. The chatbot enhances the user experience and acts as an efficient virtual assistant for the university by bridging the gap between user inquiries and institutional resources.