The application of natural language processing (NLP) in the analysis of contracts for the generation of files, particularly those processed in a notary, is revolutionizing document management in the legal sector, allowing the automatic extraction of key information, such as names, dates and specific clauses through named entity recognition (NER), and the detection of errors and ambiguities in contracts, suggesting improvements that avoid misunderstandings and disputes. The objective was to develop an application that uses models, techniques and NLP tools to analyze and identify the data of the participating entities, which are of great importance when generating the file for processing purposes. The solution includes ChatGPT to analyze and identify critical data of participating entities essential for generating files required for public deed registration. To measure the results obtained from this application, the time that the operators in charge needed to generate the files and the margin of error were considered. The results indicate that the time used to generate these files was reduced by almost 80% and the margin of error when making records was reduced by 7%. It was concluded that the application developed for the analysis of public contract registry files met its initial objective, although it could be improved to reduce the margin of error, emphasizing the importance of NLP as a key technology in the automation of complex tasks and increased productivity.