TeachMe

TeachMe Chrome Extension
Taking a screenshot w/ the extension
Chatting w/ the extension
Getting relevant topic links w/ the chat feature
Video animation playing on the right side in extension

Premise

Whatever happened to the days where the math teacher pulled out the 1's, 10's, and 100's blocks to explain how math works? Well with TeachMe, we aimed to capture that same feeling of learning- easy, understandable, and fun- and apply it to any problem, no matter how far along in your math journey you are.

What is TeachMe?

Our TeachMe algorithm starts with user input- be it in text form, image upload, or copy-pasted, and we provide an animated video tailored to the problem. Videos will go in depth to explain every step of the problem, as well as common areas that students have had problems with in the past when learning that concept. Additionally, we provide specific links to resources that could help you understand the concept even more, as well as a chat-bot to answer any lingering questions you may still have. Once all of your questions have been answered, TeachMe also gives you the option to download the video for future use, so when it comes time for exam prep or you just need a refresher, TeachMe videos can stay along with your class materials for your course.

How we built it

Our project primarily runs on Python, JavaScript, and HTML/CSS. Once the request is fulfilled, our system combines the chats and images with our own prompts, and given to 3 different agents. One agent scans the input and breaks down the problem and how to solve it, complete with the steps outline in math. At the same time, another agent is taking the same input and scouring through the internet for useful resources and links to help with your specific problem. The last agent takes the broken down solution from the previous agent and converts the text into a Manim script, mimicking a 3Blue1Brown-style of video to visualize the solution to your specific problem in a easily digestible format. Finally, the script is put through a couple of security measures to ensure the script runs, and voila! your video is ready to be played and shared.

Challenges we ran into

Manim is a generally small program, as it was released by the creator 3Blue1Brown as a open source language to try to replicate his videos. Because of this, there is not really a big sample size in a lot of agents specifically for Manim, so ensuring that we would get a running script every time was a big challenge. Prompting the agents for every step to ensure that we would get the exact style of output that we wanted was especially a challenge, but we were eventually able to overcome these obstacles.

Accomplishments that we're proud of

Going into this project, we knew that during the time constraints that we had, it would be a tall order to get everything we needed to get done for the video aspect of the app. There were a lot of times where we thought that it would be too hard to implement, or we would not be able to get what we wanted with the video, but when that video showed up embedded for the first time, you can see the difference. Its hard now to look back and imagine this app without the video aspect, and its definitely the part of the app we are the most proud of.

What we learned

Everybody likes to talk about how hard coding is, how much time you have to put for projects, and the overall work needed for a project such as this. But when you are coding something you genuinely are interested in and can see the value of, all of the "hard" coding aspects seem like a small price to pay for what you are able to make. Each of us here at TeachMe genuinely believes in the application and use for this project, and it made the long hours, energy drink-fueled weekend all worth it.

What's next for TeachMe

We aren't stopping here! We here at TeachMe genuinely believe in this project, and as we are proud of what we get to show u today, we know there is always room to grow and add new things. We believe that we could build this app into something great, and we'd all like to continue to show that this is more than just a proof of concept- its a new way to learn.