MathWhisperer

showing how to solve 3x^2 + 4x - 7 = 0 with steps
showing how to solve 4x^2 + 2x - 7 = 15 with steps

Inspiration

We thought to ourselves how could we change the world, and stumbled upon a video of a visually impaired student using an outdated technology to assist him with math. The technology would tediously ask him to input his equation one letter at a time. This was extremally inefficient and at the end of it, only gave him the answer. We knew that we would have access to Wolfram Alpha API and thought why not solve this problem by combining them. Now to get the solution the user can simply dictate the solution and has the option of receiving a step by step guide on how to solve it.

What it does

MathWhisperer transforms verbal math queries into solutions using a minimalist keyboard interface. Pressing the space bar initiates the process, allowing users to dictate equations. The system then vocally guides them through a step by step solution.

How we built it

We used voice recognition to receive the math queries, then used APIs to translate and solve the problems. The backend is built on JavaScript to process the math queries and create a solution with steps. The frontend uses React to receive auditory data to pass to the backend, and outputs the backends' solution.

Challenges we ran into

Converting the user input into a mathematical equation that Wolfram Alpha would be able to parse was a hurdle faced early into our development process. Then, we struggled to take Wolfram Alpha's step-by-step solution and translate it with ChatGPT to a concise yet effective script. A concise explanation was crucial as runtime increased rapidly as the output increased.

Accomplishments that we're proud of

We are proud of our ability to properly plan our project and then execute it as a group. We spent the first half an hour of planning the infrastructure and dividing the work. Then, we were successful in working together, handling merge conflicts and helping eachother debug. This was the first time that we had lots of conflicts, but we were able to over come them with minor set backs.

What we learned

We gained experiencing using JavaScript and React to create a website. Furthermore, the experience of using Wolfram Alpha's API and OpenAI API gained insight into different technologies. We were exposed to real world of development and basic project management.

What's next for MathWhisperer

MathWhisperer is currently limited by API runtime and acceptable input, complicating the accuracy of the program. In the future we plan finetuning the API calls to maximize the overall efficiency for a better user experience.