Traditional library management systems in academia mainly focus on organizing resources and controlling access, offering limited capabilities for intelligent knowledge retrieval and personalized learning support. While recent studies have investigated AI-based recommendation systems and chatbots for educational assistance, many current methods depend on generic information sources or lack integration with institution-specific academic content, which diminishes their reliability and educational relevance. This paper introduces Pustakalaya, an advanced framework for library management and knowledge assistance that combines structured digital library resources with a Retrieval- Augmented Generation (RAG)-based conversational interface. The proposed method creates a curated institutional knowledge base derived from digitized textbooks, reference materials, and academic documents, which facilitates semantic retrieval and context-sensitive response generation grounded in verified sources. Unlike general- purpose chatbot systems, Pustakalaya limits its retrieval to educational materials and features adaptation based on student profiles to modify the depth of explanations and recommendation strategies in accordance with individual learning abilities and academic contexts. To enhance the effectiveness of retrieval, semantic embeddings are utilized for indexing documents and matching queries, ensuring precise content selection without depending on external web searches or large-scale generative model training. Experimental testing on controlled academic datasets demonstrates that the proposed framework offers reliable, interpretable, and curriculum-aligned responses while enabling personalized access to content with lower system complexity. By merging intelligent library management with knowledge-focused conversational engagement, Pustakalaya presents a practical and viable solution for advanced academic learning environments.

 Index Terms—Library Management System, Retrieval- Augmented Generation (RAG), Natural Language Processing, Conversational AI, Semantic Embeddings, Academic Chatbot, Knowledge Information Base, Django Framework