The results of designing a software platform for intelligent text processing to support decision-making on the management of scientific research are presented. The software platform is designed for storing and processing various types of scientific and technical documentation, as well as automating the tasks of searching, comparing and analyzing documents and information. The documents being processed (scientific articles, reports and technical assignments for research) have a certain internal structure and specific features, and the list of tasks assigned to the system cannot be fully fulfilled by existing software tools. The architecture of the system is based on the modular principle for flexibility and extensibility of the project. The developed modules and components are combined into a working layout (minimal viable product), the existing and planned functionality is adjusted based on the test results. Comprehensive intellectual analysis of documentation includes both the processing of formal features of documents (deadlines, format compliance, the presence of key terms) and semantic analysis of text (annotation, comparison with assigned tasks, compliance with the subject) and to implement this, a multi- stage processing cycle is required, combining various methods of processing text data.