DEV Community

Cover image for Automating Student Reviews with AI in Pupilfirst LMS
Bodhish Thomas
Bodhish Thomas

Posted on

2

Automating Student Reviews with AI in Pupilfirst LMS

The integration of Artificial Intelligence (AI) into educational systems is a fascinating frontier. One area where AI can make a significant impact is in the review of student submissions. At Pupilfirst LMS, we've leveraged our existing Github Action integration to automate this process, making it more efficient and personalized.

Pupilfirst LMS already had a robust Github Action integration in place. This system pushes each student's submission into a new branch of a dedicated repository. From there, Github CI actions can be triggered automatically, and we have actions to grade and report the status back to the LMS. This existing setup provided the flexibility needed to build and test a new system without altering the core functionalities.

Grade Action: https://github.com/pupilfirst/grade-action
Report Action: https://github.com/pupilfirst/report-action

The main challenge was to convert student submissions, which are in JSON format, into a format that the AI model could easily understand. After several attempts, we realized that the best approach was to pass the JSON submission as-is, but with an additional explanatory prompt.

# The student's submissions will be an array of objects following the provided schema:
{
  "kind": "The type of answer - can be shortText, longText, link, files, or multiChoice",
  "title": "The question that was asked of the student",
  "result": "The student's response",
  "status": "Field for internal use; ignore this field during your review"
}
Enter fullscreen mode Exit fullscreen mode

This allowed the OpenAI model to discern which fields to consider and which to ignore.

The AI Auto Review Action was born out of this need. The logic is straightforward: we inform the AI about the context of the submission and prompt it to generate a grade and feedback. The AI model returns the data in a JSON format that is compatible with our existing grade action.

The data flow for review action

It took just a day to write the code, test the action, and deploy it. The system is now used across various courses for automated submission review.

The integration of AI into our LMS has streamlined the review process, reducing the manual workload and enhancing the learning experience with personalised feedback.

It's a testament to how Generative AI addition can revolutionize traditional educational systems.

If you want to test the Action checkout https://github.com/pupilfirst/ai-review-action

Hostinger image

Get n8n VPS hosting 3x cheaper than a cloud solution

Get fast, easy, secure n8n VPS hosting from $4.99/mo at Hostinger. Automate any workflow using a pre-installed n8n application and no-code customization.

Start now

Top comments (0)

AWS Security LIVE!

Join us for AWS Security LIVE!

Discover the future of cloud security. Tune in live for trends, tips, and solutions from AWS and AWS Partners.

Learn More

Instrument, monitor, fix: a hands-on debugging session

Join Lazar for a hands-on session where you’ll build it, break it, debug it, and fix it. You’ll set up Sentry, track errors, use Session Replay and Tracing, and leverage some good ol’ AI to find and fix issues fast.

Tune in to the full event

DEV is partnering to bring live events to the community. Join us or dismiss this billboard if you're not interested. ❤️