Large language models as a driver of research and innovation in the digital ecosystem of Bauman Moscow State Technical University

This article examines the challenges of increasing the efficiency of research and innovation activities in the context of the transition to a full innovation cycle university model. Based on an analysis of strategic documents from Bauman Moscow State Technical University, the article substantiates the need to create the "AI Korolev" scientific and educational platform as an integrator of interdisciplinary research and development in the field of artificial intelligence. The role of large language models in implementing the platform's key functions is explored: intellectual support for research processes, automation of scientific information processing, acceleration of technology transfer, and the formation of an innovation ecosystem. An architectural model for integrating large language models into the platform structure is proposed, ensuring seamless interaction between researchers, developers, industrial partners, and students. The proposed model's contribution to achieving the development program's target indicators is demonstrated, including increasing the share of research and teaching staff from 20% to 75%, increasing the volume of R&D from 5.9 billion to 30 billion rubles by 2030, and increasing the revenue of Bauman Brotherhood companies from 5 billion to 50 billion rubles by 2030.