Background: The transition to university life exposes students to significant academic, administrative, and emotional challenges, particularly during the initial phase of the academic year. In response to these challenges, the Faculty of Psychology and Educational Sciences (FPSE) at the University of Geneva established “SOS Rentrée” service, a temporary peer-to-peer support service aimed at assisting students—especially newcomers—in navigating the complexities of academic regulations and procedures. The service, operated by trained students and supervised by academic advisors, frequently involves interpreting complex and often ambiguous regulatory frameworks, a process that can be both time consuming and error-prone.

Methods: This research project aims to develop a conversational AI agent to support “SOS Rentrée” collaborators in addressing student inquiries more efficiently and accurately. The methodology involves: (a) qualitative analysis and categorization of typical questions posed during the service, based on past interactions; (b) construction of a structured knowledge base, derived from academic regulations and frequently asked questions; (c) design of a hybrid system architecture, integrating a KG (to organise regulatory information) with a pre-trained language model (to enable natural language understanding and generation).

Expected Results: The project anticipates that the AI agent will: (a) enable rapid and precise information retrieval, reducing response times for standard queries; (b) alleviate the cognitive load on collaborators, allowing them to focus on complex or sensitive cases; (c) optimise the overall efficiency of the support process by automating responses to frequent and regulation-based inquiries. Ultimately, the system seeks to enhance both the responsiveness and reliability of the “SOS Rentrée” service.

Conclusions: This initiative highlights the potential of integrating artificial intelligence into peer-support services in higher education. While automation can optimise performance, human oversight remains essential to address complex or sensitive cases, ensuring a balance between technological innovation and empathetic support. The project aligns with the broader objective of enhancing student support systems through human centred technological solutions.