Inspiration
We were inspired by some of our team members own experience being international students and peers we met at our respective colleges and how they struggled with learning content not due to ability but rather language constraints which affect understanding. They have to acquire new, complex knowledge while concurrently translating it to their native language creating large cognitive load. We wanted to build a tool that carries that load, allowing a student to mastering the actual subject matter.
What it does
Students reading dense documentation or textbooks can highlight and screenshot text. The chrome extension then injects a popup which utilizes an LLM to instantly translate the text into their native language and provide an "Explain Like I'm 5" summary. It then incorporates a feature for memory with a quiz, review guide and problem solving.
How we built it
The Chrome extension was built using vanilla JavaScript, HTML and Tailwind CSS (via CDN) to facilitate text highlighting and shadow DOM UI injection. We used Node.js and Express to build a proxy server which securely contains our LLM API key, handles CORS and routes our specific prompt structure to said LLM.
Challenges we ran into
GIven this was a first hackathon for all of us, we had initial blocks in setting up our project and ideation, which was overcome with collaboration and a degree of agentic AI assistance.
Accomplishments that we're proud of
We built an app that does not simply use AI to provide answers, but rather uses AI as an active pedagogical tool to reduce cognitive load, thereby assisting international students in their education.
What we learned
We gained a deep understanding of the differences between content scripts, background workers and popup environments in Manifest V3.
Log in or sign up for Devpost to join the conversation.