Case study
How Elon University evaluated course-grounded AI tutoring in a live rollout
Spring 2026 — Dr. Mustafa Akben deployed TUEL AI across his business courses. Elon publicly reported 80%+ voluntary adoption and 90%+ average exam scores among engaged weekly users. TUEL presents these as observed outcomes from one attributable deployment, not a universal guarantee.
- Active users
- 250+
- Voluntary adoption
- 80%+
- Avg exam score
- 90%+
- Correlation
- r = 0.32
Deployment lead
Dr. Mustafa Akben led the first TUEL deployment at Elon University.
Dr. Mustafa Akben is Assistant Professor of Management and Director of Artificial Intelligence Integration at Elon University. His public faculty profile describes research spanning deep learning, generative AI, institutional readiness, and organizational behavior, which made him a strong academic lead for a verified AI tutoring rollout.
The problem
Dr. Mustafa Akben, Assistant Professor of Management and Director of Artificial Intelligence Integrationat Elon University, wanted to give students 24/7 AI tutoring — but existing tools had a fatal flaw. Generic AI chatbots aren't grounded in course materials. Students get plausible-sounding answers with no way to verify them against the textbook. For a professor who publishes research on learning outcomes, that wasn't acceptable.
The solution
Mustafa partnered with TUEL AI to build course-specific AI tutors that are grounded in his syllabi and textbooks. Every AI response cites the exact page it came from. Students can click to verify. Faculty control AI behavior per course through model profiles and feature toggles.
- Course materials uploaded and indexed — syllabi, textbooks, lecture notes
- Every response cites specific pages students can verify
- 16 interactive learning tools beyond chat — flashcards, quizzes, concept maps, simulations
- Socratic teaching method built into the AI — guides, doesn't give answers
- FERPA-compliant with row-level tenant isolation
Deployment proof
Elon generated both academic outcomes and operational proof. The deployment story includes student adoption, faculty control, and institutional oversight instead of a single marketing metric.
- 80%+voluntary adoption
- Students chose to use the AI tutor without being required to
- 90%+avg exam score
- Closed-book exam average among engaged weekly users
- r = 0.32
- Highest individual correlation with exam performance of any study method
- 250+active users
- Spring 2026 deployment cohort
Testimonial
“What stood out to me most is that more than 80% of my students voluntarily used the Elon AI Tutor as part of their study routine. When you hand your students a thoughtfully designed AI tutor along with clearly described instructions on how to effectively use it, they utilize it as a learning aid rather than a learning shortcut.”
Timeline
2024
Concept development: Mustafa identifies need for course-grounded AI tutoring
December 2025
Elon AI Tutor launches for spring 2026 semester
Spring 2026
250+ active users, 80%+ voluntary adoption, 90%+ exam averages among engaged weekly users
Next
Multi-course expansion and research publication
Ready to review whether this rollout model fits your institution?
We'll walk through the student, faculty, and admin surfaces that matter to your evaluation and show how TUEL scopes rollout, controls, and proof.
Request a Demo