RubriqAudit

RubriqAudit is a web app (with optional mobile reviewer view) that helps schools and training programs detect inconsistent grading across instructors and cohorts. Instructors upload rubrics and anonymized samples of graded work (or export from the LMS). The system flags “grading drift” using simple statistics plus AI-assisted pattern checks: where one grader is consistently harsher on certain rubric rows, where comments don’t match scores, and where similar submissions receive very different marks. It then generates a short moderation packet: side-by-side exemplars, suggested calibration prompts, and a record of decisions for accreditation audits. This is not a magic auto-grader—humans still grade—but it reduces regrade requests, appeals, and accreditation stress by making fairness measurable and repeatable. It’s realistic for institutions that already have rubrics and want defensible assessment without rebuilding their entire assessment workflow.