Approach
Generative AI has rapidly spread across higher education, yet teachers still lack clear pedagogical frameworks for integrating it into their courses.
Research in the cognitive science of learning shows that AI can, when used as a shortcut, undermine some of the mechanisms that support learning. When integrated into a structured instructional design, however, it can instead enhance learning outcomes. CampusFrame was developed from this perspective for higher education.
Overview
CampusFrame provides the standard functions of a learning management system — programs, courses, course materials, schedules — and allows each course to be paired with an AI tutor whose pedagogical configuration is set by the instructor. The result is an AI-enhanced learning environment for students, and a tool for monitoring and analysing teaching and learning.
Using the platform requires no additional work from the instructor. The materials they upload in the ordinary course of teaching are sufficient: the platform builds a course-specific semantic index, which the tutor uses to answer questions, strictly within the boundaries thus defined.
Users
For students
From a single interface, students access their program's schedule, the materials uploaded by their instructors and, for every course, an AI tutor available at any hour. The tutor's answers rest exclusively on the course content, and every answer is traceable back to its source, a page in a PDF, a paragraph in a document, a slide, reachable in one click.
For instructors
Instructors continue to prepare their courses and upload their materials just as they already do on conventional learning platforms. The tutor uses these materials as they become available, and restricts the scope of its answers to the corpus thus assembled. In return, instructors gain a pedagogical analysis of the difficulties students encounter: recurring questions, poorly understood concepts, points at which students tend to get stuck, or materials that give rise to misunderstandings.
For program leaders
Aggregated at program level, student interactions with the AI tutors constitute a new source of observation for program leaders. They can help in better understanding the pedagogical difficulties encountered across the curriculum and provide concrete information for the academic coordination of the program, for example by making it possible to:
Pedagogical stances
Whether the request is a clarifying question, a call for practice, an exercise to be reworked, or an open-ended discussion, the appropriate pedagogical approach is not the same. The AI tutor adapts its stance to each of these situations, drawing on principles established by the learning sciences. The choice of stance can be left to the tutor, decided by the student, or set by the instructor as part of an activity designed for that purpose.
Design
Scientific foundations
The tutors’ behaviour draws on principles established by the scientific literature on learning: explicit instruction, retrieval practice, formative feedback. CampusFrame is designed to ensure that generative AI supports evidence-based teaching practices.
Founder
CampusFrame is designed and led by a program director at a French grande école, who is also a researcher in cognitive science. The platform is therefore conceived from within higher education, drawing on direct observation of the needs experienced day to day by instructors and program leaders.
Security & sovereignty
Hosted on a sovereign European cloud based in France (Scaleway). By default, no data is transferred outside the European Union. GDPR compliance will be confirmed through an independent external audit during the pilot phase.
European Cloud & AI The AI tutor can be finely configured, restricted to specific sessions, or even disabled entirely for a course. The platform retains, in every case, the functions of a complete learning environment.
Every answer the tutor gives is traceable back to its source within the course, reachable in one click. Instructors can configure the tutor so that it never answers beyond the pedagogical content provided for the course.
The architecture is designed to remain portable from one language-model provider to another (Mistral, Anthropic, Scaleway, etc.). The institution is not locked in to any single vendor.
Partner programs 2026–2027
CampusFrame is currently seeking a small number of programs willing to use the platform with their students during a free pilot phase. The only contribution expected in return is pedagogical: a critical perspective and feedback based on real use. Participating programs gain access to the platform at no cost, while their experience helps guide its development. No sales contract or commercial commitment is involved.
What the program receives
What the program commits to
White-label deployment (in time)
In time, programs that wish to may deploy the platform under their own identity — their name, their visual identity, their address. The institution remains in control of what its students see; CampusFrame is merely the publisher, in the background.
Contact
Program directors and institutional leaders interested in discovering the platform, or in considering a pilot phase, may complete this form, or write directly to contact@campusframe.ai.