GradeGuard

GradeGuard is a web app (with an optional LMS plugin) that helps instructors and academic integrity teams spot likely AI-assisted or contract-cheated submissions without pretending it can “prove” anything. It ingests assignments, rubrics, prior student writing samples (if available), and submission metadata, then produces a risk report: unusual style shifts, citation anomalies, prompt-like phrasing, version-history gaps, and rubric-mismatch patterns. It also generates an instructor-facing “next step” script for fair follow-up questions and an oral check template aligned to the assignment. The product is built around workflow: batch triage for large classes, evidence packaging for integrity cases, and policy-aligned documentation. It’s not a detector badge; it’s a decision-support system that reduces time spent on hunches and creates consistent, defensible processes for schools.