The expanding application of Artificial Intelligence (AI) has transformed multiple sectors, enabling new possibilities for automation and personalization. In the domain of reading management, many digital platforms still restrict users’ ability to create custom virtual bookshelves, thereby diminishing the overall experience. This work presents the development of Bibliware, an AI-driven reading-management system designed for flexible organization of virtual libraries. Our methodology involved building a web application that integrates the k-means clustering algorithm for book-category recommendations and the LLaMA language model for generating automatic summaries from work descriptions. As a result, Bibliware delivers a flexible solution that allows users to assemble personalized bookshelves and receive intelligent suggestions, optimizing digital library organization and facilitating the discovery of new literary works.