| Abstract [eng] |
In today‘s world of so rapid technological advancement in business, more and more challenges are being addressed through artificial intelligence and robotization. In recent years, the integration of robots into both domestic and industrial applications has become almost commonplace, especially given that robots can be physical or software. However, Robotic Process Automation (RPA), which has emerged in the business and service industry, is a relatively "young" and "promising" technology that uses solutions to automate routine work - robotizing human actions in IT systems. The automatic reading of the content of financial documents using artificial intelligence and RPA technologies is a very relevant topic in today's business world. The purpose of such RPA systems includes the detection of essential fields of a document (date, amount, supplier, vendor, etc.) and the retrieval of information. Although there are a number of commercial solutions on the market focused on solving this problem, it is still a challenge to qualitatively read and recognize all the required fields regardless of the structure of the financial document. This thesis presents a solution to the problem of reading of the content of financial documents automation using RPA (Robotic Process Automation) and artificial intelligence technologies. The result of the research work is a program that is able to recognize the specified fields efficiently and accurately, regardless of the structure of the financial document, using the most advanced technologies of artificial intelligence and RPA, thus further increasing the efficiency of the program. The introduction to the thesis paper contains a description of the problem, its relevance, the objectives pursued and the tasks to be undertaken to solve the problem. The analysis section reviews both RPA methodologies and alternative ways to automate routine work. This section also examines the products currently available on the market (software, systems, prototypes, etc.), identifying their weaknesses, strengths, application capabilities, limitations, and functionality. In the practical part of the work, there is a description of the proposed solution - the developed system starting with the specification of requirements and ending with the detailed design of the software robot. The analysis of the quality of the developed system and the possibilities for improvement are presented in the research part. An experimental study of the developed and deployed software and its improvements is presented in the experimental part. The final section of the document presents results, conclusions, and future insights. |
