The object of the research is the decision support system in the treatment of lung cancer, the subject of the research – the use of a multi-scenario interface in the construction of decision support systems. One of the problem areas in software development is the need for multi-criteria adaptation of interfaces to users. This problem became especially acute after the introduction of quarantine when various automation systems began to develop rapidly, aimed at reducing direct contact between the customer and the service provider. If earlier software users were the more or less related group, now the difference began not only at the level of technical qualifications. Now, when developing software, more attention should be paid to physiological and psychological differences between users, features of hardware and software, environment, and other criteria. In the current situation, it turned out that in most cases automated systems are used by persons who are not interested in these systems but simply have to use them. One of the options for solving this problem is to create an adaptive universal interface. This research is aimed at analyzing methods for implementing multi-scenario decision support systems in the treatment of lung cancer. In the research, attention is paid to the following aspects: adaptive intelligent interface, architecture and structure of the adaptive intelligent interface, algorithms for the functioning of agents of adaptive system interfaces. In the research, the system was used by 500 participants for 30 days. The benchmark was the type of data display scenario selected at the start and end of the day. The research showed a gradual transition of users to scenarios of higher complexity, which involve the analysis of all available information. The tendency of reverse transitions decreases with time, and from the 18th day of using the system, the type of the selected interface changes in rare moments. These results proved the possibility of using automatically configurable interfaces, and bringing them to the final form will be achieved in 18–20 days of using the system.

The object of research is the electronic learning system. The subject of the research is the method of using microservices in the construction of online systems. One of the most problematic areas in the development of high-load online systems is the coordination of all microservices in a single system and the distribution of the load on hardware resources at critical indicators of system utilization. This leads to the complication of the process of development, implementation and operation of the training system, as well as high requirements for the personnel who will support the operation of the system. In the research, during the transition from the monolithic architecture of the e-learning system to the microservice architecture, the main indicators of the server hardware and the average response time to user requests were monitored. These indicators were fundamental when setting up the system as a whole and balancing the load during its operation. The proposed method for the implementation of the system can significantly reduce the hardware requirements and reduce the response time of the system under high load conditions (from 10,000 unique users per unit of time). Also, this method greatly simplifies the development and modification of online systems that use a large number of different user roles and differentiation of levels of access to the system. The obtained results of the approbation of the method allow to consider it an effective tool for the development of online learning systems with multivariate access to educational materials. Unlike existing monolithic architects, the proposed method allows to manage system resources and apply new settings without rebooting, which allows to ensure the continuity of system operation. As a justification for this method, options for the implementation of online training systems and load balancing settings are proposed. The management of load balancing in the microservice architecture of the implementation of online systems is based on the analysis of the load indicators of processor cores and the use of RAM by system services.