SUMMARY
The subjects of the research are models and methods of optimization and automation of personnel selection in safety-oriented systems with their further formation into project teams on the example of civil protection service, law enforcement agencies of national police, as well as the Armed Forces of Ukraine. Purpose: to develop models for the formation of project teams in safety-oriented systems, taking into account the parameters of personal human qualities and professional skills of candidates for competitive selection based on the global experience of companies in the field of human resources management using software information expert systems. The following tasks are solved in the article: to identify the most effective modern models that can be applied in the civil protection system, taking into account the specifics of the service, on the basis of the use of the European project management toolkit and programs in the field of personnel, organizations, firms and companies. It is necessary to expand the terminological base of management of projects, programs and project portfolios by introducing a new definition of a safety-oriented system. For this research, popular modern automated recruitment systems, as well as expert information systems models, would be considered to capture all relevant information from the database and graphically describe the process of forming project teams in safety systems and to generalize the candidate index model. The following methods are used: human resource management using automated HRM systems, expert information systems and index numbers. The following results were obtained: the model of the structure of the information expert system and the process of knowledge accumulation through the software were developed, an index method for formation of project teams (on the example of military formations) in safety-oriented systems was proposed. Conclusions: a generalized model of expert information system was developed for the selection of personnel into safety-oriented systems based on an index method.