Chatflow

Chatflow utilizes AI agents to provide User interface that allows users to interact with various systems using natural language. Our sophisticated engine accurately understands user intent and dynamically displays the appropriate UI, optimizing the user experience.

This approach enables users to effortlessly navigate and engage with complex websites or products. Instead of the traditional point-and-click method, users can access multiple functionalities and perform tasks through a conversational interface, making the process both intuitive and efficient.

You can try it out here

Watch this video to learn more.

Join our Discord to know what's going on in development and to ask questions to the maintainers about the project and how to contribute: https://discord.gg/fJ5ecMmsSf

RAG architecture

Running the App

Before running the app, please install Docker first

Set up the environment

React UI

• cd chat-ui
• npm install
• npm run build
• npm start

Backend

• Start the databases

  docker-compose up -d redis postgres video-chat-server

• Install the dependencies

  pip install -r requirements-dev.txt

• Navigate to the backend src cod

  cd server/src

• Create a .env file and set all required variables

  cp server/src/.env.template server/src/.env

• Replace the OpenAI API key with your own key in the .env file

• Start the backend service locally

  # Load the initial data
  python load_data.py
  # Start the server
  python server.py

• Log into the app http://localhost:3000/assets#/login with the following credentials

  username: [email protected]
  password: 123

• Visit http://localhost:8880/api/docs to see the API docs

• Access the Redis Vector DB UI on http://localhost:8001/redis-stack/browser

Local Development with Docker

• Build the docker image docker build -t apssouza/chatux:latest .
• Set OPENAI_API_KEY_GPT4 and OPENAI_API_KEY_GPT3 environment variables
• Run docker-compose up to start the app
• Visit http://localhost:8880/api/docs to see the API docs

Github demo

• Go to Github and create a new classic token https://github.com/settings/tokens
• Set the Github Token in the local storage of the browser

localStorage.setItem("external-token", "github-token")
localStorage.setItem("external-url", "github.com")

• Prompt the Github related actions. Ex List Github repositories and display as a chart with the fields: full_name and forks_count

Backlog

Please look at the issues for the backlog

Leave a star if you like the project