KSU presented the concept of the University LLM.eco corporate artificial intelligence ecosystem.

University LLM.eco – a corporate artificial intelligence ecosystem for the management and development of the university. Its concept involves creating a single intellectual center that unites various educational, scientific, and administrative processes. The authors and developers – the assistant to the rector on strategic issues, Maksym Poltoratskyi, and the programmer of the Digital Expertise Center Dmytro Volyk.

The ecosystem is built on the principle of functional distribution. It includes five main areas: e-Academic — academic activity, e-Governance – administrative management, e-Knowledge — knowledge management and information retrieval, e-Quality – quality assurance of education, and e-Engagement — communication.

At the same time, all these areas operate within a common digital ecosystem, which includes a corporate knowledge base and KSU24. This approach is intended to help avoid situations where each department uses a separate AI assistant, while information and work logic remain fragmented.

An important element of the system is the e-Orchestrator – a central mechanism that organizes interaction between functional agents. The user formulates a request in a single chat, after which the system analyzes the task and determines which specific agents need to be involved to execute it. Thus, the interaction based on the "each with each" principle is replaced by a managed request processing route.

During the presentation, it was also shown how such a model can work in practice: from a user request through orchestration and the involvement of necessary tools – to the formation of a final result, specifically an analytical report or a document.

A separate part of the session was dedicated to practical work with AI assistants. Participants reviewed examples of queries, discussed how to correctly formulate tasks for artificial intelligence and how to act if the system produces an error or the result does not meet expectations.

Course materials on KSU-online: https://ksuonline.kspu.edu/course/view.php?id=7750

KSU Press Center