Learn It

Project 3 submission for the General Assembly Software Engineering Immersive.

Team members:

• Brendan Fox
• David Kuperman
• Michael Zhao
• Shiam Dulla

Introduction

Learn It is a web application that leverages the power of OpenAI's ChatGPT API to automatically generate interactive quizzes on any topic supplied by the user. The application provides a seamless experience for users to create, save, browse, and delete quizzes, as well as share them with friends.

List of features

1. Quiz Generation: Users can input a topic of their choice, and the application uses the OpenAI ChatGPT API to automatically generate a custom quiz with multiple-choice questions.
2. Quiz Management: Users can save quizzes they generate, browse their saved quizzes, and delete quizzes they no longer need.
3. Share Quizzes: Users can easily share quizzes with their friends through a generated link, allowing others to take the quiz.

Technologies used

1. Express.js web server run on Node.js
2. Postgres database
3. Bootstrap CSS
4. AJV JavaScript library for JSON object validation
5. JavaScript front end Single Page App

Architecture

The application follows a client-server architecture, with the frontend and backend communicating through API endpoints. The backend server is built using Express.js and connects to a Postgres database for storing and retrieving quizzes. The frontend is built as a SPA using JavaScript, with Bootstrap CSS for styling. The OpenAI ChatGPT API is integrated into the backend server to generate quizzes based on user input topics.

Below is the image of the basic architecture of the app including front-end, back-end and integration with OpenAI chatGPT API:

Below is a diagram showing the db schema:

Architecture

https://sei-project-3-learn-it.fly.dev/signup.html

Future features

1. Persist Dall-e images in AWS S3 bucket
2. Persist app information through fly io postgres database
3. Personalized Quizzes: using user feedback on questions they got wrong to influence the generation of future quizzes, such as asking questions focused on areas that the user has struggled with, providing a more personalized learning experience.
4. Quiz Customization: Allowing users to further customize the generated quizzes e.g. by specifying the number of questions.
5. Quiz Statistics: Providing users with additional statistics on their quiz performance, such as time taken.
6. User strengths / weaknesses: investigate the use of ChatGPT to provide a summary of a users strengths / weaknesses.
7. Unit tests: As a starting point, could be provided on the functions calculating user statistics. Ultimately, would ideally have complete test coverage.
8. How to guides / FAQ: provide some advice on ways that users can get the most out of the app, e.g. students could use the 'context' field to paste in text from a textbook, so that the quiz is more relevant to what they are studying.