Early access · engineering courses

Grading that shows its work.

Partial Credit reads handwritten and typed engineering homework, suggests rubric items with the page evidence behind them, checks the math exactly where it can, and never releases a grade until you approve it.

Students: join a course with your code, or open Partial Credit from your LMS course.

FERPA compliant. How we meet each requirement

Question 2 · pages 3–4

2 State feedback 8 / 10

0Correct gain derivation

−2Sign slip in the characteristic polynomialsuggested

−4Controllability not justified

Exact check: closed-loop poles match the target

Awaiting your approval

Built with subject-specific engines for redundancy and higher accuracy.

Built for how engineering and math are graded

Six stages from assignment to gradebook.

  1. Step 1

    Set up

    Upload the assignment. Bring your key and rubric, or let AI write both.

  2. Step 2

    Collect

    Students upload a PDF or phone photos and match pages to questions.

  3. Step 3

    Grade

    AI suggests rubric items and cites the work behind each one.

  4. Step 4

    Verify

    Exact math checks recompute the answers on questions you pick.

  5. Step 5

    Review

    You accept or change each suggestion and approve every grade.

  6. Step 6

    Publish

    Students see grades once you publish. Totals go to your linked LMS.

Grading view: a handwritten pole-placement page beside the rubric for that question, with three rubric items applied for 3 of 3 points.
Each question opens on the pages that answer it, next to its rubric.
Page mapping: thumbnails of a three-page handwritten submission, each tagged with the questions it answers.
Pages are matched from question headings or your handout’s layout; fix any in a click.
  • Due and late-due times
  • Rosters and grades synced with your LMS
  • Extensions and regrades
  • A dated log of every decision

Step by step → All features → Screens show synthetic coursework.

Most AI graders only read the answer. Partial Credit also checks the math.

On questions where you turn on a check, Partial Credit recomputes the student’s answer with exact math instead of comparing it to how your solution key is written.

  • Correct, but written differently. Two valid state-space models can use different matrices; the check confirms they describe the same system.
  • More than one right design. It confirms the student’s gain places the requested poles, rather than matching yours.
  • Says so when it can’t tell. Pass, fail or inconclusive, shown next to the suggestion. It never changes a score; you do.
Deterministic checks panel: controllability matrix, rank, open-loop and desired polynomials and the unique gain K = [6 3], each marked as holding.

Measured, not promised. Try it on grades you already gave.

Import a past assignment you have already graded. See exactly where Partial Credit agrees with you, where it is more lenient, and which rubric items it misses, before it touches a live course.

Request early access
  • Where it agrees with you
  • Where it is more lenient than you
  • Which rubric items it misses

Works inside your LMS

Partial Credit works with Canvas, Moodle, Brightspace and Google Classroom.

  • Canvas
  • Moodle
  • D2L Brightspace
  • Google Classroom

How the LMS integration works →

  • Link once. Link your LMS course or Google Classroom class to your Partial Credit course. For Canvas, Moodle and Brightspace, your administrator first registers Partial Credit with a one-time link.
  • Rosters stay current. Students, TAs and instructors follow LMS enrollment, synced every day and whenever you open the course from the LMS. People you added yourself are never changed.
  • Grades flow back. When you publish, approved grades go to the LMS column you choose. Scores changed in the LMS gradebook are kept unless you choose to overwrite them.
  • Students click in. Students open their assignments straight from the LMS course or the Classroom assignment.

FERPA

FERPA compliant.

Homework and grades are education records. Partial Credit is built so your institution stays in control of them. See each FERPA requirement and how we meet it.

  • Only the people in your course. Students see only their own work and published grades. TAs grade; only instructors publish.
  • Used only to grade. Student work is never used to train AI models, and no AI runs on it unless your course staff start it.
  • Deletion you control. Permanently delete any submission. Erasing everything an instructor holds is available on request.
  • A trail you can audit. Grading decisions, grade publications and sign-in security events are logged.

Principles

See it on your own course.

Early access is open to engineering instructors and TAs. Bring one assignment you have already graded.

Students: join a course with your code.