Extrakce strukturovaných dat z PDF souborů s využitím LLM a integrace s ERP systémem

Abstract

The content of this bachelor thesis is the design and implementation of a system for the automated processing PDF invoices, primarily using of Large Language Models (LLM). The implemented application allows users to extract important data from invoices sent via email attachments, perform semantic matching of the identified data, create necessary records in ERP system modules, and upload the source PDF files to attached network file storage. The theoretical part of the thesis includes a research of possible approaches to solving this problem and a description of the specific technologies selected for the development of the application. The practical part focuses on the design and implementation of the application, specifically using the n8n automation platform.