Software Development on Automation of Legal Function in the Sphere of Property-rental Relations with the Help of Language Models and Python Programming Language

The graduate qualification work consists of two projects: 1) A program for processing EGRN extracts; 2) Chat-bot of the legal assistant. These projects were created to solve specific problems in the sphere of property and lease relations. The reason for the creation of these projects is a direct interest, demand on the part of business. That is, these projects were initially created to solve real, applied tasks. The combination of the program for processing extracts from the EGRN and legal chatbot is an advanced tool designed to accelerate the work of the legal department and increase the efficiency of each individual lawyer. The EGRN extracts processing program allows to automate the processing of PDF-documents, which significantly reduces the time required to analyze the actual information on the real estate object (whether it is changes in cadastral value, area, or other characteristics). At the same time, automatic processing of a large amount of data significantly reduces the probability of errors that could occur during manual data processing. And the result of the program in the form of Excel-table sorted in chronological order for each real estate object will also allow you to identify possible errors and inaccuracies that may be contained in the state register for specific objects. The program for processing EGRN extracts is written using the Python programming language. The final realization of the project is an executable exe-file. Chat-bot legal assistant is able to adjust to the user's requests and has an internal “memory” of the existing dialog with the user. That is, over time the bot becomes more and more personalized, and therefore more suitable for solving urgent problems of a lawyer. At the same time, the bot itself is trained on current legislation, as well as relevant legal theory. That is, the bot provides its answers to legal questions based not only on specific laws and legal rules, but also on the underlying principles of law. The chatbot legal assistant is implemented using large language models and no-code tools. The final realization of the project is a telegram bot. Thus, within the framework of the Graduation Qualification Work, all the knowledge gained during the study was applied in practice, which is aimed at the realization of real, non-theoretical business projects: 1) writing code in a programming language; 2) interaction with neural network models; 3) working with no-code platforms.