MarkInMinutes

MarkInMinutes is an AP history grader tool. An AI grading assistant for teachers and lecturers that reads student submissions against a rubric, cites evidence for each score, and returns structured feedback with a human review step built in.

MarkInMinutes is an AI grading assistant built for secondary school teachers and university lecturers who need to work through large stacks of student submissions without cutting corners on feedback. The product handles the laborious part of assessment: reading each submission against the rubric, locating supporting evidence, and writing structured commentary. The educator keeps the part that requires professional judgment, which is deciding what the final grade should be. The workflow is deliberately simple. An instructor sets up an assignment, attaching instructions and a grading rubric in whatever format already exists, including PDF, DOCX, or a rubric built inside the app. Student work is then uploaded as PDF, DOCX, XLSX, PPTX, or code files, with batch upload supported so an entire cohort can be submitted in one pass. A multi-agent pipeline takes over: one stage extracts evidence for each rubric criterion, another scores dimensions against calibrated benchmarks, a third audits scores for consistency and hallucination, and a fourth drafts improvement coaching. Most submissions return in five to ten minutes, and results include an executive summary, dimension scores, per-error feedback, and citations that let the instructor verify the reasoning before signing off. The tool suits humanities essays, business reports, computer science assignments, and mixed media submissions where charts, tables, diagrams, and source code all need to be read together. Its stated alignment with expert human graders is above 90 percent, which positions it as a first-pass reviewer rather than an autograder. Compared with general purpose chatbots that will score a single essay when prompted, MarkInMinutes is built around class-sized volume, rubric calibration, and a documented audit trail. The vendor is explicit about scope: audio and video submissions are out of scope, Excel formula evaluation is not yet available, and instructors are expected to review and adjust every result.