Utilization of Large Language Models in the Educational Process

Abstract

This paper addresses the potential of Large Language Models (LLMs) to support students in the educational process by providing personalized assistance, explanations, and examples. LLMs are used by many students, but they often lack up-to-date information or specific knowledge on the subjects they need. This paper looks at how using Retrieval-Augmented Generation (RAG) can help by linking language models to external sources of information so answers can be more accurate and relevant. The proposed solution is implemented using the LangChain framework, which simplifies integration with vector databases, embedding models, and custom prompt workflows. The system is designed as a modular architecture with specialized agents that respond to different types of student queries. This approach demonstrates how LLMs can be enhanced to better learning experience.