Unstructured AI use is already part of student study habits
Ignoring that reality does not help instructors understand when AI support is useful, misleading, or pedagogically harmful.
Why this project exists
Students are already using general-purpose AI tools while studying. TutorBot explores a different model: course-grounded AI support shaped by faculty intent, explicit assignment structure, and research guardrails.
Faculty need a way to shape the help students receive
TutorBot lets instructors define assistant behavior, provide question context, and decide what kinds of support are appropriate.
Research matters as much as access
The platform is intended to study structured AI tutoring in real educational settings, not simply to offer another productivity tool.
How TutorBot works
The platform is built around faculty-authored assignments, guided student interactions, and analysis of the resulting tutoring sessions.
1
Faculty configure assistants in course context
Instructors set assignment structure, author question context, and define how assistants should behave for tutoring or feedback.
2
Students engage with structured support
Students encounter AI help through assignment-aware workflows instead of a free-form chat disconnected from course intent.
3
Researchers and faculty review the interaction patterns
TutorBot captures the conversational context needed to examine how structured AI assistance is actually being used.
What TutorBot is not
These guardrails describe the limits and intended use of the project.
Not a grading engine
TutorBot is not intended to replace course assessment or make grading decisions on behalf of instructors.
Not a generic chatbot portal
The project focuses on assignment-aware, faculty-shaped support rather than broad unrestricted conversation.
Not part of a student academic record
TutorBot is not meant to be FERPA-regulated recordkeeping and should not become part of a student's academic record.
What faculty can shape
TutorBot is intentionally designed to keep pedagogical control with instructors rather than with a default model prompt.
Assistant role and tone
Faculty can configure assistants for tutoring, feedback, or other educator-defined roles that fit a course workflow.
Question-level context
Assignments include question text, example answers, concepts, and common mistakes so responses are grounded in course intent.
Interaction boundaries
The platform supports course-specific constraints around progression, feedback flows, and the level of help students should receive.
Current project status
Pilot phaseTutorBot is being developed and piloted through Rochester Institute of Technology as a research platform for structured AI integration in coursework.
- The current public experience is intended to explain the project and its educational framing.
- Faculty and academic staff can request access when participation is appropriate for their role or pilot context.
Developed through
AI Foundry
Rochester Institute of Technology
Explore the project
If you are evaluating TutorBot for teaching or research use, start with the project background and guardrails. If you already have access, you can sign in directly.