Inspiration

Coming from South Texas, two of the team members saw ESL (English as a Second Language) students being denied of a proper education. Our team created a tool to break down language barriers that traditionally perpetuate socioeconomic cycles of poverty by providing detailed explanations about word problems using ChatGPT. Traditionally, people from this group would not have access to tutoring or 1-on-1 support, and this website is meant to rectify this glaring issue.

What it does

The website takes in a photo as input, and it uses optical character recognition to get the text from the problem. Then, it uses ChatGPT to generate a step-by-step explanation for each problem, and this output is tailored to the grade level and language of the student, enabling students from various backgrounds to get assistance they are often denied of.

How we built it

We coded the backend in Python with two parts: OCR and ChatGPT API implementation. Moreover, we considered the multiple parameters, such as grade and language, that we could implement in our code and eventually query ChatGPT with to make the result as helpful as possible. On the other side of the stack, we coded it in React with TypeScript to be as simple and intuitive as possible. It has two sections that clearly show what it is outputting and what ChatGPT has generated to assist the student.

Challenges we ran into

During the development of our product we often ran into struggles with deciding the optimal way to apply different APIs and learning how to implement them, many of which we ended up not using or changing our applications for, such as the IBM API. Through this process, we had to change our high level plan for the backend functions and consequently reimplement our frontend user interface to fit the operations. This also provided a compounding challenge of having to reestablish and discuss new ideas while communicating as a team.

Accomplishments that we're proud of

We are proud of the website layout. Personally the team is very fond of the color and the arrangement of the site’s elements. Another thing that we are proud of is just that we have something working, albeit jankily. This was our first hackathon, so we were proud to be able to contribute to the hackathon in some form.

What we learned

One invaluable skill we developed through this project was learning more about the unique plethora of APIs available and how we can integrate and combine them to create new revolutionary products that can help people in everyday life. We not only developed our technical skills, including git familiarity and web development, but we also developed our ability to communicate our ideas as a team and gain the confidence and creativity to create and carry out an idea from thought to production.

What's next for Homework Helper

As part of our mission to increase education accessibility and combat common socioeconomic barriers, we hope to use Homework Helper to not only translate and minimize the language barrier, but to also help those with visual and auditory disabilities. Some functions we hope to implement include having text-to-speech and speech-to-text features, and producing video solutions along with text answers.

Share this project:

Updates

Submission history