GRAS

What it does

Our project is a tool for students that allows them to reflect on their work, such as essays, using an AI tool that checks their work against a provided rubric.

How we built it

GRAS uses the OpenAI API for GPT4-turbo and Vision to process the rubric into plain text, then it uses the GPT4-turbo API to analyze the user's essay and provide them with a percentage and some feedback on how to make their assignment better aligned to the rubric.

Challenges we ran into

A primary challenge we ran into was configuring our backend to properly interface with the OpenAI API properly, as it had a lot of room for error and needed careful prompting to get optimized results.

Accomplishments that we're proud of

We are proud of our intuitive UI that allows users to simply upload an image and paste their assignment to have the API and our program to process their grade.

What we learned

We learned a lot about using GPT models and troubleshooting them.

What's next for GRAS

GRAS is currently quite simple, a future update would include more versatility for file types, a more nature oriented UI, and more detailed results.

Built With

• figma
• openai
• python
• restapi