Our project aims to develop a structured method to support teachers in deploying conversational robots such as “assistants,” thereby enriching the learning experience for students and revitalizing teaching methods. After implementing a handmade version of such an assistant in an information technology course at a business school, we want to repeat the experience in a more structured way in two other courses in different disciplines: human resources and professional accounting. Our ultimate goal is to design, document, and disseminate a structured approach to facilitate such deployments in college and university educational institutions. The conversational assistants we have in mind will be configured to answer students' questions on the specific subject matter of the targeted courses. This innovative educational tool will allow students to interact with a virtual assistant, accessible 24/7, capable of providing accurate and relevant answers.

All interactions with the assistant will be recorded, anonymized, and analyzed. This analysis will be used for the following purposes: 1- Understanding the community's interactions with the assistant; 2- Evaluating the quality of queries; 3- Identifying queries that generate the best responses; 4- Identify topics requiring further understanding; 5- Validate the design of a standardized approach to creating assistants that are easy for any teacher to deploy.
The results of this project could facilitate the gradual integration of GenAI into teaching approaches. More broadly, the configuration of the assistants and the analysis of interaction data will serve as models for the creation and use of virtual assistants in other academic disciplines and training contexts.