The PRET19 project aims to carry out a comprehensive analysis of the nineteenth-century loan registers of several Parisian university libraries. These historical documents provide invaluable insights into the circulation of books and the intellectual engagement of the academic community during a transformative period. By reconstructing the relationships and trends between borrowers, the project offers a unique perspective on the intellectual landscape of Parisian universities. In this first phase, the registers come from three different university libraries and exhibit a great diversity in layout, handwriting and content, which poses significant challenges for data processing. To address these challenges, we developed a document processing workflow that effectively combines automatic handwritten text recognition (HTR) with manual processing and validation. In addition, we provide a detailed description of the database, which is designed to comprehensively model the information extracted from the registries, ensuring that the data is structured in a way that adequately responds to anticipated research requests. Once completed, the processed data will be from winter 2024 accessible via a dedicated website, providing a comprehensive digital resource for historical research.