AutoMark Summer 2024 Update: Adaptive grading, inline comments, and easier export

Julian Meyer
Julian MeyerJuly 1, 2024 Product Update

Inline comments screenshot

In January 2024, my brother Holden and I had an exciting realization about automated grading: AI should handle the tedious tasks.

Our goal was to empower teachers without taking over the entire grading process.

Grading papers is often a slow, monotonous, and repetitive task, leading teachers to assign fewer papers each year. This means students get fewer opportunities for practice, and sometimes it takes over four weeks to get feedback due to teachers' heavy workloads. With teachers already overburdened, students end up getting even less practice. By enabling teachers to assign more practice homework, AI can significantly enhance student learning outcomes.

Today, we're thrilled to announce a major update to our grading system:

  • 💬 Inline comments: AI-suggested comments on essays. When grading an essay, the AI will now comment on specific phrases in the student's work and provide feedback based on the rubric. Teachers can easily approve or edit any of these comments or add their own.
  • 🔮 Adaptive grading: As you grade more submissions, the AI learns your feedback style and what to look for in the assignment. Teachers can adjust feedback, and the AI will suggest similar feedback for future submissions.
  • 🦉 PDF export: You can now export assignments as PDFs to send to each student or print. This includes inline comments, scores, and feedback.

We hope these updates make your grading process smoother and more efficient!

💬 Inline comments

One piece of feedback we've received numerous times asks for the ability to comment inline on students work, similar to Google Docs. We've now added the ability to add comments to students' work.

Inline comments take the grading process to the next level. Instead of just overall feedback, students now receive feedback on specific phrases in their essay.

Suggested comments

To make it easier for teachers to leave feedback on the essay, the AI will first suggest comments to add on the essay before assigning a score and feedback. This huge advance in automated grading will allow students to get feedback quicker and iterate more on their essays, leading to better outcomes.

🔮 Adaptive grading

"Show don't tell" is common writing advice, and that's the approach we took when building our new feedback system. Previously, teachers would have to write custom instructions and configure the output with dropdowns, but now the auto-grader learns your style over time.

Instead of writing instructions on how the assignment should be graded, just grade a few submissions. AutoMark will transparently learn your style of grading.

🦉 PDF export improvements and CSV export

One of our main goals is saving teacher's time. This really boils down to having the best grading experience possible. One area we realized could use improvement is getting grades out of AutoMark.

We already support seamless import/export through LMS systems when integrating with districts, but we want to make it easy to import/export for everyone on our individual plan.

You can now export a an Excel or CSV file of all the grades on the assignment, including question-level and rubric-level scores for each student.

🦉 Student Feedback PDF

Finally, with inline comments, we know it will be important to return feedback to the student, so we added a PDF export option where you can export a personalized feedback PDF for each student.

PDF export screenshot

You can also customize what's included in the PDFs to include rubric scores, overall feedback, and/or inline comments.

Tell us what you think!

What's next for AutoMark? We're working on improvements to rubric extraction, import/export improvements, better onboarding and more.

We hope these updates will help give significantly better and more timely feedback with AutoMark.

We're giving the tool away for a limited time, so sign up now and let us know what you think! You can always email us if you have any suggestions or feedback!