discussion or speech involving multiple participants, through multiple forms of communication
A figma-like canvas editor to visualize and engage in branching conversations across various LLMs. Each prompt & completion is represented as a node, and different models can be selected for each node.
- Enable multimodal capabilities such as images
- Enable real-time collaborative editing
- Allow resizing nodes
- Explore templating conversations for different use cases
- Next.js & tailwind
- Flask
- Redis
- Firestore
- Langchain
- Together.ai
- GCP
This app is deployed using Google App Engine. Deploy backend first.
- Configure
backend/app.yaml
and add any required env variables. Referencebackend/app.example.yaml
. - Run
gcloud app deploy
inbackend/
.- Run
gcloud app logs tail -s backend
to debug issues.
- Run
Deploy frontend second.
- Configure
frontend/app.yaml
and add any required env variables. Referencefrontend/app.example.yaml
. - Run
NEXT_PUBLIC_BACKEND_ROOT_URL=<APP_ENGINE_BACKEND_URL> yarn build
. - Run
gcloud app deploy
infrontend/
.- Run
gcloud app logs tail -s frontend
to debug issues.
- Run