Alloy

Sample notes drawn on our app
A student using the AI summarization tool to explain a term
Student notes being drawn on ipad
A collaborative doc featuring python code snippets

Inspiration

The lack of a comprehensive notetaking app. All the good features of notetaking apps are either locked behind paywalls or split across different apps, with no single one featuring every feature we wanted. So, we set out to make an app that solved that issue we faced.

What it does

Note taking app with all the best features of every note taking app. Features rich formatting, code snippets, live ipad drawing integration, transcription, AI summarization threads, and live collaboration.

How we built it

We used the following tech stack: next.js + typescript Tailwind CSS Convex Database Clerk Auth tldraw for ipad draw integration Rich Text: tiptap

We began by first having one member develop a base such that the rest of our team could build features on top. Working in parallel, 2 of our group members created component mock ups in Figma. Once there was a solid base which the team could realistically build off of asynchronously, we began to implement the components previously made in Figma. It was at this point our fourth team member arrived allowing us to work at maximum capacity. We had backend development, bug fixes, and UI updates working in parallel to create our project: Alloy.

Challenges we ran into

Our transcriptions originally were very slow, and on top of that they weren't even totally accurate to what the user would be saying. It became quite discouraging, and we were even considering scrapping the entire transcribe option entirely. After much deliberation as a team, we decided that the ai integration of speech to text was integral to our project success. To overcome our previous hurdles, we feed all transcribed information into the Groq api to organize and punctuate thoughts, which ultimately made the entire process of transcription faster too. Through collaboration and teamwork, we managed to deliver a high quality project without needing to cut any of the integral parts.

Accomplishments that we're proud of

transcription and voice to text
live collaboration
AI note summarization
organized documents

What we learned

Throughout this hackathon, we had to learn many libraries from the ground up. These included AI integration into our project (Explain with AI powered by Gemini), database management (Convex), frontend component development & design via Figma, and several other technologies used in our project.