AI Student Matching

What it solves

When students hand in work on paper, the tedious part isn't grading, it's sorting 40 scanned PDFs into the right student's folder before you can grade anything. AI Student Matching automates that sorting step.

Before you start: enrolled students need roll numbers

AI Matching identifies each scanned submission by the roll number (or name) it finds written on the page and matches it against your course roster. Make sure your students are enrolled in the course with roll numbers set (see Roster and Manage Enrollment) before you upload, without a roster to match against, the AI has nothing to match to.

Uploading and matching

1. Open the AI Matching tile on the course dashboard.
2. Select up to 30 scanned PDF files at a time (one file per student submission is typical, but multi-page PDFs are fine).
3. Upload. The AI model reads each PDF, extracts the roll number or name it finds, and matches it to the corresponding enrolled student.
4. Review the results. Any files that couldn't be matched (roll number or name not found, or not recognizable) are listed separately so you can handle them manually, either re-scan them more clearly or use Submit on Behalf in ZenGrader to file them yourself.
5. Open a matched PDF directly in ZenGrader with one click to start grading right away.

Recent AI Matches

The same screen keeps a running list of your Recent AI Matches (Last 30), which file matched which student, when, and a direct Open PDF in ZenGrader link, so you can pick up a batch-matching session exactly where you left off.

A quick note on Canvas imports

If your scans already come out of a Canvas LMS export as a single zip of files named by student, ProfDesk also offers a straightforward zip import that matches by filename (roll number or name) without needing the AI reader, useful when your file names are already clean and structured.

How much you can match, and what to do beyond that

AI Matching draws from your credits according to the batch’s actual model token usage (see AI Grading for your monthly credit numbers), and on Free it also draws on a separate paper allowance:

• Free, 10 scanned papers a month matched to your roster by AI.
• ProfDesk Pro, Unlimited AI matching: scan a whole class at once, up to 30 papers per batch.

Re-matching a single file inside a batch you have already run does not use another paper, and neither does re-importing a corrected scan of a paper you already brought in.