Inspiration
We were inspired by the ChatGPT-conversation-based AI, where users can interact with the AI through texts. Based on that, we did some brainstorming about possible products that could make use of AI to provide information for users. The need for reducing time in reading the whole document and having an assistant that is available 24/7 has led our team to the idea of having AI as a Teaching Assistant.
What it does
Our project's main purpose is to help users summarize their readings and answer users' questions based on the uploaded PDF. Our main goal was to minimize the time used for reading a whole document to understand the overall scope of the reading. Also, if users ran into problems related to the reading, they can ask our AI bot to generate some relevant answers.
How we built it
We built the front end first with HTML, CSS, and JavaScript. For the backend, we imported OpenAI and handled logic in Python.
Challenges we ran into
We faced some challenges with the development of the website such as how to connect the front-end and backend together, or how to parse the uploaded PDF to OpenAI so that it can read and generate a summary for the documents. Lastly, we had some difficulties in providing a quick-responsive website that can provide the summary right away once the button is clicked.
Accomplishments that we're proud of
We were proud of our progress at the end. The product achieved our primary goal, which is to provide a summary and QA services using AI although the process is not as effective as expected. However, within the allowed time, we are proud to say that we tried our best in this and our product can function as desired.
What we learned
We learned a lot from this experience such as teamwork skills, how to implement OpenAI to help us read a document and generate a summary, or some of our teammates learned about how Github works, how to code in Python, how to link backend and front-end, etc. Above all, this hackathon was a great chance for us to explore what we want to try and experience the new thing.
What's next for Artificial Intelligence Teaching Assistant
We still need to make the website respond quicker, handle a heavier PDF file and handle multiple requests concurrently. The current process involves uploading a file to appwrite server and then downloading it for processing. The purpose of this was to allow user to revisit their summaries at a later time but could not be done due to time constraints.
Log in or sign up for Devpost to join the conversation.