Expanding Student Service Capabilities Using AI: Implementing LLM and Rag Into University Information Retrieval Systems

Traditional digital technologies in higher education, including official websites and PDF documents, are often fragmented, limited by language, and do not allow for efficient data retrieval, particularly for international students. This paper presents an approach for integrating Large Language Models (LLMs) and the Retrieval Augmented Generation (RAG) method into existing university services. The proposed architecture provides semantic search, real-time data updates, multi-language support, and interaction with existing backend applications without the need for model fine-tuning. A chatbot prototype was developed, allowing students to access information in their preferred language based on up-to-date university data. Testing demonstrated high accuracy in information delivery and a seamless user experience. The proposed approach carries practical significance.