AI-Driven Academic Assistant: An Intelligent Chatbot System Leveraging Llama-3 and Rag for Enhanced University Support Services

In today's higher education environment, the demand for cost-effective student services is rising due to increasing student numbers and the density of academic data. This paper outlines the development and implementation of a chatbot system for a world-class university, offering intelligent, personalized real-time assistance to students. The system processes various document formats-PDFs, images, and Word files-using PyMuPDF (+ Poppler), Tesseract, Pillow, and python-docx. Powered by the Llama-3 (Large Language Model Meta AI) model for natural language understanding and generation, the chatbot provides accurate responses to complex inquiries. The system uses FastAPI for backend operations, Langchain for advanced query identification, Chroma for high-performance vector storage, and ChatGroq for smooth natural language generation. Additionally, tools like pandas, openpyxl, and chromadb enhance overall system performance. This research shows that a chatbot-driven approach can efficiently track student queries, providing personalized, timely responses while reducing administrative workloads and significantly improving student satisfaction. The system aligns with global initiatives, such as UNESCO's (United Nations Educational, Scientific and Cultural Organization) AI framework, supporting inclusive and culturally sensitive applications. This work lays the foundation for AI-driven academic support systems, contributing to near-term and long-term advancements in educational AI deployment.