EnlightenAI's AI scores each student submission against your rubric and writes personalized feedback. It uses your rubric plus your graded training samples as the model of what "good grading" looks like, so the scores and comments match how you grade. Results stream in live, and you review and edit everything before students see it.
What does the AI actually produce?
For each submission, the AI generates:
A score for every rubric criterion, using your criterion levels and point values.
Comprehensive feedback written in student-friendly language.
A short area of strength and area of growth (15 words or less each).
A score classification of HIGH, MEDIUM, or LOW based on the total.
How does the AI know how to grade like me?
The AI is given your full rubric (criteria, descriptions, levels, and points) along with your custom prompt details, such as grade level, the task students were given, and any reference text. It also pulls in your most recent graded examples for that assignment as training samples, modeling its scores and comments on what you actually did. The more samples you grade, the closer the AI matches your style.
What does "responses stream in" mean?
Feedback is generated live and appears on screen as it's written, rather than all at once after a wait. EnlightenAI's parser reads the response as it streams, so you can watch feedback take shape and the system stays resilient even if a response is interrupted.
Do students see AI feedback automatically?
No. The AI drafts the grade, but you stay in control. After grading runs, you can review and edit any score or comment before releasing it. Feedback reaches students only once you finish grading, so nothing is published without your review.
Still need help? Contact our support team.