
Today, educational institutions compete not only through their academic programs but also through the quality of digital interaction. Learning platforms handle dozens of inquiries daily about courses, events, deadlines, and enrollment requirements. When a prospective participant searches for a relevant program, they expect an immediate answer — without having to browse dozens of pages. This is why educational institutions are increasingly implementing RAG agents — intelligent assistants that combine the capabilities of large language models with the precision of knowledge base retrieval. Such solutions relieve managers of routine workload, provide users with instant answers, and significantly increase the conversion of inquiries into registrations, as potential customers receive all the necessary information in just a few simple steps. kmbs has naturally incorporated this logic of personalized user interaction into its own digital initiatives.The business school systematically cultivates a new management culture, operating at the intersection of ideas, leadership, and innovation. The key to this lies in its flagship MBA programs, as well as programs that set the development trajectory for those aspiring to reach the MBA level in the future. Since 2024, the school has expanded this trajectory by launching a new direction — artificial intelligence for executives — helping leaders scale the impact of technology within their organizations. Innovative thinking is embedded at the core of the kmbs approach — from strategic programs to formats that offer leaders new perspectives on decision-making. This is reflected in their international module at POLIMI Graduate School of Management, where participants explore industry trends; in innovative programs for commanders of the Defense Forces of Ukraine; and in events dedicated to business analytics and strategic management. For kmbs, innovation is not merely a technological attribute but a component of its educational philosophy. It is therefore natural that the school integrates intelligent tools into its audience engagement, choosing solutions that reflect its vision for the future of education.
Guided by its vision and business needs, the client turned to SMART business to develop its own RAG agent, “kmbs Assistant” — an intelligent assistant built on Microsoft Copilot Studio that helps website visitors navigate the business school’s programs, events, news, and content more efficiently. Want to learn how the solution was implemented? We’re sharing the details in our new success story!
For kmbs, the website is the primary channel through which prospective participants explore programs and submit enrollment inquiries. While partner companies systematically send their teams for training, the influx of new participants is driven primarily by the website.
The scale of interaction is impressive: program pages alone attract between 30,000 and 60,000 sessions each month, creating high expectations for the speed, accuracy, and effectiveness of communication with prospective participants, partners, and corporate clients.
Despite such high audience activity, prior to launching the RAG agent the business school lacked a centralized tool that could automatically help users navigate the kmbs website. All inquiries were handled manually — through forms, email, and hotline calls. This approach placed a significant burden on managers and did not always provide the immediacy modern users expect.
Moreover, the kmbs team clearly recognized challenges associated with the website’s internal search. Analytics showed that users often failed to find the necessary information on their first attempt and, instead of continuing to interact with the site, resorted to phone calls or written inquiries.
The idea of creating an intelligent assistant emerged about five years ago, long before the implementation of the RAG agent by SMART business. At that time, the concept involved a conventional chatbot without artificial intelligence — essentially a “button-based menu,” where each response had to be manually programmed. This approach required enormous effort to maintain and update and did not deliver the desired level of flexibility. Ultimately, it became clear that this was not the tool that could truly support the audience, and the idea was abandoned.
The situation changed when AI-powered tools became more familiar, accessible, and sufficiently mature for use in real business scenarios.
We returned to the concept of a digital assistant with a new objective — to create a solution that thinks in context, works with the business school’s knowledge base, and helps users find the communication path that resonates with their professional ambitions. For example, if a prospective participant is searching for marketing programs, the intelligent agent should not only answer questions but also compile a relevant selection of programs starting soon. And if a question falls outside a typical request, it should be transferred to a manager without interrupting the conversation. As a result, SMART business proposed a solution that precisely matched our needs — from knowledge base functionality to integration with internal processes.
It was extremely important for us that the RAG agent always provide up-to-date and verified information. Program start dates, pricing, descriptions, faculty lists, and topics are constantly updated, so users must receive accurate data in real time. It was critical that the bot respond to inquiries while synchronously integrating updates from our website, ensuring the most convenient and transparent experience possible for prospective participants.
The implementation of the RAG agent on the kmbs website was carried out in stages, enabling the gradual configuration of all system components and ensuring the accuracy and reliability of the intelligent assistant for business school users.
Each step — from infrastructure preparation to website integration and testing — was carefully executed by the SMART business team to guarantee the desired outcome.
At the initial stage, SMART business specialists configured the Microsoft Azure environment, which became the foundation for deploying the RAG agent. This included ensuring system stability, security, and scalability — all critical for handling a high volume of user inquiries.
The next step involved a detailed analysis of the existing kmbs website content and file structure. Based on this, the SMART business team created the knowledge base for the RAG agent — a centralized repository from which the agent retrieves up-to-date information about programs, events, faculty, news, and other business school content. Importantly, this knowledge base is not static: the integrator configured an automated parser that continuously updates it with new materials from the kmbs website. As a result, the agent always operates with the latest information, while systematized and structured data ensures accurate and relevant responses for users.
At this stage, the development team created the agent description and added the kmbs website as the primary knowledge source, logically divided into sections and categories. Special attention was given to configuring the “Escalation” topic, responsible for automatically forwarding complex or atypical inquiries to program managers, ensuring continuous communication with users and personalized responses to their needs.
To integrate the RAG agent with the website, an API connection was configured to automatically transmit and receive information between systems. This enabled the “kmbs Assistant” to interact with users directly within the website interface.,The next step involved testing the solution together with the kmbs team, verifying response accuracy, the proper functioning of the “Escalation” topic, the agent’s response speed, and the usability of the interface.

For kmbs, audience engagement has always been built on trust and natural communication. The school does not have a traditional sales department, as each manager is responsible for their own educational program, and processes are designed to avoid a formal “sales vibe” — the atmosphere of intrusive commercial pressure, where client communication is reduced to meeting KPIs or quickly closing deals. Instead, a kmbs manager acts more as a consultant and partner in choosing an educational path: helping individuals determine whether a program truly aligns with their professional needs and only then guiding them toward participation.
A kmbs manager’s role extends far beyond simply “enrolling a client in a program.” Managers conduct a brief mini-interview to understand whether the program genuinely meets the participant’s needs. This stage is critical for the future group dynamic: if, among 20 participants, several people are not a good fit, the entire group is affected. Therefore, the mission of the “kmbs Assistant” is to guide a person toward communication with a manager and program director — not to replace this stage with a chatbot.
Considering the above, the client is exploring the possibility that, in the future, all inquiries submitted through the chat assistant will be automatically duplicated in the CRM system. CRM integration would allow managers to obtain a complete view of inquiries and work with them without additional manual steps, while the “kmbs Assistant” would remain a tool that guides users toward communication, completing an application form, scheduling a call, or arranging a meeting with a business school representative.
Another idea under discussion is training the bot to gather more information about users — where they work, what they do, and their professional interests. This functionality would enable managers to receive a ready-made summary even before the first conversation with a prospective candidate, reducing the time required to clarify essential details.
At the same time, the kmbs team emphasizes that the balance between informativeness and user comfort must be preserved so that people do not abandon the dialogue due to excessive questions or a sense of pressure. The assistant’s task is to collect context delicately without turning the conversation into a questionnaire — keeping users engaged while maintaining a natural communication experience. And with the solution from SMART business, this is entirely achievable.
Even if a company has strict KPIs or specific business requirements, RAG agents from SMART business can adapt to virtually any interaction scenario. They can automatically collect information, conduct personalized dialogues, route inquiries to the appropriate managers, or integrate with CRM — everything depends on the organization’s needs. SMART business develops intelligent assistants tailored to each company’s unique characteristics, ensuring the technology supports business goals and enhances the experience for all users.
The final stage included presenting the completed solution to the kmbs team and gathering feedback for further improvements and deeper customization of the “kmbs Assistant” to meet the client’s specific needs.,SMART business also provides administrative support, ensuring the proper operation of the RAG agent and enabling future expansion of its functionality — such as adding new sections, programs, knowledge sources, or implementing additional integrations with other business school systems and tools.,Within just a few days, the client team and SMART business specialists completed the full cycle of work — from infrastructure setup and knowledge base creation to testing and presenting the finished RAG agent — ensuring it was fully ready for live interaction with users.