Tutor, Not Solver: Designing a Guardrailed AI Assistant for Learning in Higher Education: A Design Case of PeteChat

Abstract

Generative artificial intelligence (AI) tutors hold significant promise for higher education, yet designing systems that scaffold learning without undermining academic integrity remains an open design challenge. This paper presents PeteChat, a course-aligned AI tutor developed and piloted at a large U.S. research university and documented through the lens of design-based research (DBR). Drawing on formative expert evaluation with teaching assistants and user-experience and developer stakeholders, we report eight transferable design principles for assessment-aware AI tutors, ranging from homework guardrails and debugging scaffolds to self-regulated learning support and instructor-facing customization tools. The system is built on a locally hosted large language model from the Llama-3 family, enhanced with retrieval-augmented generation (RAG) grounded in course-specific materials. Rather than reporting controlled experimental outcomes, this design case foregrounds the situated design reasoning, iterative refinement, and principled decision-making that shaped PeteChat across four development phases. The principles and methodological approach offer actionable guidance for institutions seeking to deploy responsible, integrity-preserving AI tutors at scale.