8pm Sunday. Paper 23 of 40. Your comments got shorter an hour ago.
Good feedback takes six minutes a paper. You have forty papers and one evening. That math beats you every single week, and your students can tell which third of the stack their paper was in.
Yours forever. Delivered the second you pay.
Three teachers this was built for
Raise your hand if one of these is you.
The Sunday griever
Forty papers land every week and your Sunday night goes with them. You have done the math on six minutes a paper and the math always wins.
The comment-shrinker
Paper 3 gets two thoughtful paragraphs. Paper 30 gets "good point, expand." You know your students can tell, because you could tell when you were one.
The careful skeptic
You want the time back but you refuse to let a machine grade your students. Good. That rule is built into every prompt in here: the AI never assigns a grade.
It was not built for anyone hoping to hand the stack to ChatGPT and go watch the game. Every rating and every sentence that ships is still your call, on purpose.
Who is selling you this
I am a professor of entrepreneurship at John Carroll University and the Founding Director of the Bill & Sue Donnelly School of Leadership and Social Innovation. I cofounded the Experiential Entrepreneurship Curriculum, used at about 200 colleges. More than 20,000 teachers read my newsletter, How to Teach With AI.
I grade real stacks every week, and this system came out of my own Sunday nights, tested on sample stacks until it caught things I had missed: a stats misconception at the exact step it entered, a number that did not follow from the student's own data. Four of those catches are listed further down this page.
The stack always wins.
Two thoughtful paragraphs on paper 3. “Good point, expand this” by paper 30. You know exactly when the drop-off happened, because you felt it happen.
The math is the whole problem. Six minutes of real feedback times forty papers is four hours you do not have, so the feedback shrinks to fit the evening.
Caring less to finish faster has a name: surrender on a schedule. You have been running it every Sunday, and it has never once felt like a system.
Three passes. You keep the pen.
Pass 1 drafts feedback on every submission, quoting the student's actual words as evidence. Generic praise is banned inside the prompt itself.
Pass 2 checks each paper against your rubric, proposes a rating, and marks every close call BORDERLINE instead of guessing. Pass 3 compresses your decisions into a short note that sounds like you.
You confirm, you override, you decide. The AI never assigns a grade.
Four catches from the sample stacks it was tested on
- The exact step where a stats misconception entered.
- An intermediate number that did not follow from the student's own data.
- A discussion post that read AI-written, with the specific tell named.
- A personal disclosure in a discussion thread. It refused to draft that reply. That one stays yours. It should.
Tonight's first run
- 1
Open your current stack and pick one paper.
- 2
Strip the name off and paste Prompt 0, the starter.
- 3
Read what comes back. Keep what is right, cut what is not.
- 4
Decide whether the other 39 get the same treatment.
Ten minutes, no new software. It runs in the ChatGPT, Claude, Gemini, or Copilot you already use.
What is inside
Fifteen prompts written out in full, organized by assignment type, plus the routine that strings them together.
Essays and written analysis
Feedback draft, rubric check, and voice pass tuned for essays. Comments quote the student's own sentences as evidence, and every close call gets flagged for you.
Problem sets
The quantitative versions. They trace where the first wrong number entered and whether everything after it follows from the student's own work.
Discussion posts, batch mode
A whole board in one run, with a batch cap that keeps replies short and a rule that every reply names the student's actual point, so no two replies are interchangeable.
Presentations
Slide decks and recorded talks, checked against your rubric with the same three-pass routine.
Project and lab reports
The long-form version. Structure and evidence checked section by section, so the 15-page report gets the same rigor as the 2-page essay.
Prompt 0 and the batch-session setup
The starter prompt for tonight's first paper, and the session setup that carries your rubric and standards across a full stack.
Two worked examples
A sophomore essay graded start to finish in four minutes, every decision shown, so you can see the system before you run it.
What this will not do
It will not grade for you. It proposes, flags, and drafts; every rating and every sentence that ships is your call.
It will not fix a broken rubric. The rubric check is exactly as good as the rubric you paste into it.
And the paper-3 version of you stays the ceiling. The system's whole job is getting paper 38 up to it.
Questions people ask before buying
How do I get it?
Checkout runs through Stripe. The second you pay, you land on the system itself, a page you bookmark and keep forever. Lose the link and one email to doan@betterteacherwithai.com gets it back.
What do I need?
Any chat AI you already use: ChatGPT, Claude, Gemini, or Copilot, free accounts included. Nothing to install. Prompt 0 is designed to run on one real paper tonight.
What about student privacy?
Names, IDs, grades, and anything about accommodations never go into a chatbot. The routine starts with stripping identifiers, codes instead of names, and the key stays on your machine.
Will it work for my assignment types?
The fifteen prompts cover essays, problem sets, discussion boards, presentations, and project or lab reports. If you grade something written, numeric, or presented, there is a prompt tuned for it.
Why $29?
It is the cheapest thing on this site because it should be the first thing you try. If it gives you back even one Sunday evening, you have made the trade of the semester.
Fifteen prompts across five assignment types, the three-pass routine, Prompt 0 for tonight, and two worked examples. Yours forever, on every stack you grade from here on.
Get the Sunday Night Grading SystemSecure checkout through Stripe. You get it immediately after payment.
This is also included in Better Teacher with AI Pro, along with the app, the workflow hub, and every other system I have built, from $29 a month.
See the Pro tiersYour next stack lands Sunday
Same size as always. Four hours the old way. The system is built to get it to about ninety minutes, with better comments on paper 38 than the four-hour version ever left.
Whether paper 38 gets the 8pm version of you again is now a choice.
One paper tonight. Prompt 0. Ten minutes.
Get it for $29