Syllabot, Your Academic Sidekick.

Inspiration

The inspiration for our project, Syllabot, comes from our experience as students having to dig through pages of syllabus text only to miss a crucial deadline or grading rule. We saw an opportunity to create a tool that would empower OSU students and help them succeed in their academic journey.

What it Does

Syllabot is an academic chatbot designed to simplify the academic experience for OSU students. It takes course syllabi and files to answer any questions students have. Whether it's clarifying grading rules, highlighting an unusual due date, or offering guidance on assignments, Syllabot is your academic sidekick, ready to assist 24/7.

How we built it

We built our application using a multitude of services/frameworks/tools:

React.js for the client frontend
Tailwind for styling
DaisyUI for animations and UI components
ExpressJS for our REST.API
MongoDB for the document and course-info storage
LangChain to parse PDF documents
Pinecone as the VectorDB for our vectorized document data
OpenAI's API for the generative chats
Render for frontend deployment

Challenges We Ran Into

Managing many different services and getting them to work together in fluidity.
Reading a lot of documentation -- this was the first time any of us had use Vector Databases and embeddings. It took a lot of research and heads-down coding to get it working.
Finding time to work as a team -- we all have busy lives as college students, and finding time to work for 72 hours brought some difficulties.
Depricated/incompatible packages for our authentication manager in the back end.

Accomplishments that we're proud of

Finishing our project and getting it working! We were honestly surprised at the progress we made each day and are super proud of the end product.
Learning a ton of new technologies

What we learned

Used Vector Databases for the first time
Learned concepts regarding embeddings,
Worked together as a team for the first time

What's next for Syllabot

Improve AI chat generation using more metadata to query the vector database.
Add options to remove and edit course files.
Implement global state management to enhance security.
Allow users to submit files without logging in.

🚀 Contributors 🚀

Imgyeong Lee

Oliver Elliott

Tom Nyuma

Sankalp Patil

Built With

chatgpt
daisyui
express.js
langchain
mongodb
pinecone
react
tailwind

Submitted to

Hack Education
- Winner First Place

Created by

I worked on the front-end for the landing page and chat page, and middleware for connecting to the mongoDB database. This was my first time working with Clerk, file uploads, and daisy UI.

Sankalp S Patil
hi :)
I designed the UI/UX and worked on the frontend for all pages, especially chat and manage pages with React, tailwindCSS, and DaisyUI. I designed the initial structure of our MongoDB database. I produced a utility function with Langchain, and fetched data from our backend to the frontend. Learned a lot about the workflow of the AI chat application and VectorDB.

Imgyeong Lee
I worked on the landing page, making pages responsive to data, user authentication, and the system prompt for the GPT api. This was my first time working with tailwind, clerk, and openAI's APIs. It was a very fun experience and I learned a lot of skills I'll use going forward!

Oliver Elliott
I worked on the backend API service, connecting the frontend to the backend, and integrating the Pinecone vector database, OpenAI embeddings, and Langchain VectorDBQA. Learned a lot about writing A.I applications.

T Nyuma
always enjoy a great hack! ❤️