Maestro-Prof

AI Grading Platform

Workspace
Back to Documentation

Professor Guide

Complete guide for professors using Maestro-TA to create exams, manage submissions, and grade student work.

Getting Started

Accessing Your Account

  1. Navigate to the Maestro-TA login page
  2. Enter your professor credentials
  3. You'll be redirected to the Professor Dashboard

Professor Dashboard

The dashboard is your central hub for managing exams and submissions.

Dashboard Overview

  • Statistics Cards: Total Exams, Total Submissions, Pending Grading, Graded
  • Recent Exams: Your 5 most recent exams with quick access
  • Recent Submissions: Your 5 most recent submissions with status

Creating Exams

Step-by-Step: Creating a New Exam

  1. From the dashboard, click "New Exam" button (top right)
  2. Fill in exam details: Title, Course ID, Description (optional)
  3. Upload answer key PDF (can be typed or handwritten)
  4. Set grading preferences (if available)
  5. Click "Create Exam" to save

Answer Key Format Tips

  • ✅ Use clear, legible handwriting
  • ✅ Well-organized layout with numbered questions
  • ✅ Sufficient spacing between questions
  • ⚠️ Avoid extremely small text or poor contrast

Grading Submissions

Automatic Grading Process

Maestro-TA uses Rashomon Ensemble Grading - multiple AI engines grade independently for accuracy and consistency.

  1. Find the submission in Grading page
  2. Click "Grade" button on pending submission
  3. Wait for processing (1-3 minutes)
  4. Click "View" to see results

Understanding Results

When viewing a graded submission, you'll see overall score, per-question breakdown, Rashomon metrics (variance, confidence intervals), and detailed feedback.

Understanding Rashomon Ensemble Grading

Maestro-TA uses multiple AI engines (like having multiple TAs grade the same paper) to ensure:

  • Consistency: Multiple perspectives reduce bias
  • Accuracy: Disagreements highlight ambiguous cases
  • Reliability: Score intervals show confidence levels

Score Interpretation:

  • • High Agreement: All engines gave similar scores → High confidence
  • • Low Agreement: Engines disagreed → Review manually
  • • Score Interval: Range of scores from different engines

Best Practices

Creating Exams

  • • Use clear, unambiguous answers
  • • Include multiple acceptable formats
  • • Specify partial credit rules
  • • Use descriptive titles

Grading Submissions

  • • Review high-variance cases
  • • Use feedback effectively
  • • Monitor system performance
  • • Adjust answer keys if needed

Frequently Asked Questions

How accurate is the AI grading?

Maestro-TA uses multiple AI engines for high accuracy. Score variance metrics help you identify cases that need manual review.

Can I edit grades after they're assigned?

Yes, you can manually adjust any AI-generated grades and add comments.

What if OCR misreads handwriting?

You can review and correct OCR results. The system learns from corrections.

Need More Help?

For detailed instructions, troubleshooting, and more FAQs, see the full guide or contact your system administrator.

View all documentation →