Stud.AI

Stud.AI Welcome Page
Interactive Log-in Page
Interactive Sign-up Page
Well-designed UI along with functioning study note(s) prompt for receiving user inputs
AI generated practice set given the user's study notes
User home page displaying previous queries as well as the option to add more
Functional search system to locate previous saved entries
Display page for AI generated practice sets with the option for downloading, returning to the home page and logging out

Inspiration

Our team's fascination with the vast possibilities of AI motivated us to explore its potential impact. As first-year university students grappling with a demanding workload, we understood the value that could come out of leveraging AI to enhance the educational journey. With this project, our goal was to create a tool that would simplify the studying process by providing closely tailored and thought-provoking practice questions on demand, catering to the unique needs of students.

What it does

Stud.AI is an innovative app that transforms a user's notes into personalized practice tests. Users can engage with these tests to reinforce their understanding of the material and effectively prepare for exams. Leveraging the Cohere API, we incorporated Large Language Models (LLMs) to generate practice sets specifically tailored to the inputted notes, ensuring a highly customized and efficient learning experience.

How we built it

We employed a technology stack that combined React.js, Redux.js, and TailwindCSS for the frontend. The backend, powered by Node.js and Express.js, interacted with a MongoDB database hosted on Google Cloud servers. Integrating the Cohere API for natural language processing allowed us to effectively correlate practice sets with the input notes.

Challenges we ran into

The most challenging aspect of this project was properly integrating the Cohere API calls and resposne into the React environement, as none of us had interfaced with this API before. After encountering bugs with various approaches, we persevered and found a configuration that led to success. Additionally, connecting the backend to the frontend, along with implementing secure authentication and protected routes, presented additional hurdles that demanded careful attention.

Accomplishments that we're proud of

We are proud to have created Stud.AI, an innovative tool that addresses a common issue faced by students. We are also proud of creating a full-stack web application complete with user authentication with JSON web tokens. This allowed us to create a personalized dashboard for all users where they can view their previously generated tests, similar to many popular LLMs models such as ChatGPT. This in turn would help with easing the burden that many students face in university and high school. We were also satisfied with the overall styling and look of the final product given the short timeframe of this hackathon.

What we learned

Our journey with Stud.AI equipped us with the skills to effectively utilize the Cohere API for creating innovative applications with large pre-trained LLMs. We also gained a deeper understanding of full-stack development and creating a positive user experience through UI/UX design. Finally, we came out of this project with a deeper understanding of proper backend authentication integration as well as some niche aspects of styling such as using TailwindCSS to accelerate the styling process and using 3js and Framer to create 3D animations.

What's next for Stud.AI

Looking ahead, our team is eager to introduce new features, including flash cards and test score tracking, to further enhance the learning experience. Continuous refinement of the language learning model is also on the horizon, with the goal of making generated tests even more accurate and personalised. A final exciting prospect is the addition of an image interpretation mode, allowing users to scan their notes and tests directly into the test generator for an enhanced and seamless generation process.