In year 2020, the world faced an epidemic of the new coronavirus SARS-CoV-2, that lead to a global crisis. COVID-19 crisis is the complex and challenging problem of the real world so to solve such a crisis is a good task for super intelligent system because the devil is in the details so the agent has to model the world with some really good degree of competency. We propose a conceptual scheme to solve this task by building a modeling (including using an agent-based approach) and analyzing system involving intelligent systems (artificial or hybrid). Conceptually this system has to be able to discover the most accurate model of the problem and to offer the best actions to solve it. In this article we propose the approach to solve COVID-19 problem and to build super intelligent system for this. We analyze the world experience in applying information technology to solving problems of system analysis of such a degree of complexity. A series of simple experiments to illustrate the idea proposed in the article is shown.