Edumentor: A Multi-Agent AI Study Assistant for Personalized Learning

Abstract

The rapid growth of digital education platforms has increased the demand for intelligent systems capable of delivering personalized learning experiences. Traditional online learning systems often provide static course structures that fail to adapt to individual student needs, learning styles, and knowledge levels. This paper presents Edumentor – Multi-Agent AI Study Assistant, an intelligent educational platform designed to provide personalized learning support using a multi-agent architecture powered by large language models. The system analyzes student learning requirements and dynamically generates customized learning roadmaps, recommended learning resources, practice quizzes, and interactive tutoring. In addition, the platform incorporates Retrieval-Augmented Generation (RAG) to enable document-based learning, allowing students to upload study materials and receive context-aware explanations. The system is implemented using a modular multi-agent framework where specialized agents perform tasks such as student analysis, curriculum planning, resource discovery, quiz generation, and tutoring. A vector database is used to store semantic embeddings of uploaded documents, enabling efficient similarity-based retrieval during question answering. The platform integrates modern AI technologies with an intuitive web interface to create a flexible and adaptive learning environment. The results demonstrate that the proposed system can effectively assist learners by providing structured learning guidance, improving access to educational resources, and supporting interactive knowledge exploration.

Introduction

The text discusses the limitations of traditional online learning platforms, which often follow static structures and fail to adapt to individual student needs, learning styles, and goals. This lack of personalization makes it difficult for learners to effectively engage with content and manage their own learning process.

With advancements in artificial intelligence, particularly large language models (LLMs) and Retrieval-Augmented Generation (RAG), new opportunities have emerged for creating intelligent and adaptive educational systems. These technologies enable systems to generate explanations, recommend resources, and provide context-aware answers using both internal models and external learning materials.

To address these challenges, the proposed system, Edumentor – Multi-Agent AI Study Assistant, introduces a personalized learning platform based on a multi-agent architecture. Different AI agents handle tasks such as analyzing student profiles, generating customized learning roadmaps, recommending resources, creating quizzes, and offering tutoring support. Additionally, the RAG module allows students to upload their own study materials and receive tailored, context-based explanations.

The system integrates these components into a unified platform with an interactive interface, enabling adaptive and flexible learning. Overall, Edumentor enhances the learning experience by combining personalization, intelligent assistance, and document-based interaction, making education more efficient and student-centered.

Conclusion

This paper presented Edumentor – Multi-Agent AI Study Assistant, an intelligent learning platform designed to provide personalized educational support using a multi-agent architecture. The system analyzes student inputs such as topic, knowledge level, learning goals, and preferred learning style to generate customized learning roadmaps, recommended resources, quizzes, and tutoring assistance. The platform integrates multiple specialized AI agents that collaborate to perform different educational tasks, enabling an adaptive and interactive learning experience. In addition, the system incorporates a Retrieval-Augmented Generation mechanism that allows students to upload study materials and receive context-aware answers based on those documents. The results demonstrate that the proposed system successfully combines generative AI capabilities with document retrieval techniques to deliver personalized learning assistance. The integration of a multi-agent framework improves task specialization and enables the system to handle multiple learning functions within a single platform. Overall, the proposed system highlights the potential of AI-powered educational assistants in improving accessibility, personalization, and efficiency in modern digital learning environments

References

[1] T. B. Brown et al., “Language Models are Few-Shot Learners,” in Advances in Neural Information Processing Systems, 2020. [2] A. Vaswani et al., “Attention Is All You Need,” in Proceedings of the 31st International Conference on Neural Information Processing Systems, 2017. [3] P. Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” in Advances in Neural Information Processing Systems, 2020. [4] J. Devlin, M. Chang, K. Lee, and K. Toutanova, “BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,” in Proceedings of NAACL-HLT, 2019. [5] N. Reimers and I. Gurevych, “Sentence-BERT: Sentence Embeddings using Siamese BERT Networks,” in Proceedings of EMNLP, 2019. [6] M. Honnibal and I. Montani, “spaCy 2: Natural Language Understanding with Bloom Embeddings,” Explosion AI, 2017. [7] J. Wu et al., “A Survey on Intelligent Tutoring Systems and Educational AI,” IEEE Transactions on Learning Technologies, vol. 13, no. 4, pp. 764–777, 2020. [8] H. Chen, Y. Liu, and Z. Wang, “Artificial Intelligence in Education: A Review of Intelligent Tutoring Systems,” IEEE Access, vol. 8, pp. 75264–75278, 2020. [9] J. Guo, M. Chen, and Y. Zhang, “Personalized Learning Systems Based on Artificial Intelligence: A Survey,” IEEE Access, vol. 9, pp. 123245–123257, 2021. [10] R. Mihalcea, C. Corley, and C. Strapparava, “Corpus-Based and Knowledge-Based Measures of Text Semantic Similarity,” in Proceedings of AAAI, 2006.

Copyright

Copyright © 2026 Prof. P. Anjaiah, A. Vaishnavi, G. Srivani , S. Sreeya, CH. Soumya. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.