- Live audio streaming to an arbitrary number of users through WebRTC
- Live transcription running on client through a WASM'd version of (distilled) whisper when audio streaming
- Leverage LLM's to answer student questions (output is provided in a thread of the question)
- Automatically find the part of the transcript a student's question is referring to
- Automatically attempt to answer students question
- Uses open-mixtral-8x7b
- Collaborative real-time whiteboard with media upload and panning ability
- Live messaging and reactions
- npm (to install pnpm) https://docs.npmjs.com/downloading-and-installing-node-js-and-npm
- pnpm (package manager)
npm install -g pnpm
- bun (runtime for backend service) https://bun.sh/docs/installation
- docker https://www.docker.com/get-started/
docker-compose -f docker-compose.dev.yaml up db --build
cd services/db
pnpm migrate:generate && pnpm migrate:run && pnpm seed:emoji
docker-compose -f docker-compose.dev.yaml up fireside --build
cd frontend
pnpm install
pnpm run dev
docker-compose up