This paper aims to research the technologies and the approach to implementation of AI tool for the analysis and storage of unstructured data. The technologies which will be used in the practical part, as well as researched in this paper are as follows: Java programming language, Eclipse Scout framework, PostgreSQL database and PG Admin database administrator. The tool being implemented in the application is Receipt & Invoice OCR, developed by the company Taggun. The external libraries used in the practical part of the paper are: JDBC, Gson and okHttp. The external libraries mentioned were necessary for connection to the database and the implementation of the AI OCR tool. In the first part, the purpose and the goal of the paper is explained, after that, the functional requirements for the development of the application and the implementation of the tool. The second part aims to analyze the technologies used and the theoretical terms which are related to the AI OCR tool. The last part of the paper focuses on the practical part of the implementation of the AI tools for recognizing and structuring data from unstructured sources and the theoretical part of the external libraries used. The complete application code is available from the public repository on GitHub and can be found at the link: https://github.com/dinobaresic/Zavrsni_rad_IT_2024.