Pocket Professor

For this year's theme, our group has identified an issue common among students receiving graded assignments from their professors. Very often, the solutions contain little to no feedback and explanation on what the student did wrong or why the real solution is the right solution. To tackle this problem, we designed a web application where students can submit their assignments in images and generate explanations on “Why is my answer incorrect, why is the correct answer correct?” We use the React framework for our website, and we implemented an image-to-text algorithm that parses the images, we use OpenAI API to automatically parse the paragraphs into question, my answer, and the correct answer. Then, those prompts are inputted into the OpenAI API again to generate an appropriate response. For this project, we learned how to implement OpenAI API in Python and React. Our group members each had experience in React, NLP, and HTML, but integrating those parts into a complete project was the most challenging part.

Built With

• flask
• gpt
• node.js
• python
• react