AI Assistant

A streaming AI assistant application integrating Chainlit and RAGFlow with Redis message queuing.

Complete Report

• Email: Please read：[https://alidocs.dingtalk.com/i/nodes/1DKw2zgV2PAo2rA7uLr72orN8B5r9YAn?utm_scene=team_space]

For detailed deployment instructions, please refer to the following document:

• Deployment Guide

Overview

This project provides an interactive AI assistant with streaming responses, document references, and human agent escalation capabilities. The application uses Chainlit for the frontend interface and RAGFlow for backend knowledge retrieval and response generation.

Features

• Real-time streaming responses
• Document reference integration
• Human agent escalation
• Redis-based message queuing
• OAuth authentication
• Structured logging
• Environment-based configuration

Requirements

• Python 3.12+
• Redis server
• RAGFlow API access
• Rocket.Chat server (for human escalation)

Installation

1. Clone this repository

   git clone https://github.com/iamtornado/AI-Assistant.git

2. Create and activate a virtual environment: Linux:

   python -m venv .venv

   Windows:

   python -m venv .venv

   Activate the virtual environment: Linux:

   source .venv/bin/activate

   Windows:

   .venv\Scripts\activate

3. Install dependencies:

   pip install -U fastapi uvicorn chainlit redis rocketchat_API

4. Create a .env file with required environment variables (see Configuration section)

Configuration

Required environment variables:

• RAGFLOW_API_KEY: API key for RAGFlow access
• RAGFLOW_BASE_URL: Base URL for RAGFlow API
• RAGFLOW_ASSISTANT_NAME: Name of the RAGFlow assistant
• REDIS_HOST: Redis server hostname
• REDIS_PORT: Redis server port
• REDIS_PASSWORD: Redis authentication password
• ROCKETCHAT_SERVER_URL: URL for Rocket.Chat server
• ROCKETCHAT_WEBHOOK_TOKEN: Token for Rocket.Chat webhook authentication
• LDAP_PASSWORD: Password for LDAP authentication
• WEEKDAY_USERS: Comma-separated list of Rocket.Chat users for human escalation
• OAUTH_KEYCLOAK_BASE_URL: Base URL for Keycloak OAuth server
• OAUTH_KEYCLOAK_REALM: Keycloak realm name
• OAUTH_KEYCLOAK_CLIENT_ID: Keycloak client ID
• OAUTH_KEYCLOAK_CLIENT_SECRET: Keycloak client secret
• OAUTH_KEYCLOAK_NAME: Keycloak OAuth provider name
• CHAINLIT_AUTH_SECRET: Secret for Chainlit authentication
• IT_ENVIRONMENT: Environment type (dev or test)
• REDIS_DB_INDEX_DEV: Redis database index for development environment
• REDIS_DB_INDEX_TEST: Redis database index for test environment
• LOG_LEVEL: Logging level (DEBUG, INFO, WARNING, ERROR, CRITICAL)
• PROPAGATE_LOGS: Whether to propagate logs to parent loggers (True/False)

Usage

Start the application with:

    uvicorn webhook_server:app --host 0.0.0.0 --port 8000

Project Structure

• chainlit_ragflow_streaming.py: Main application entry point
• ragflow_client.py: RAGFlow API client
• message_queue.py: Redis queue implementation
• logger_config.py: Logging configuration
• webhook_server.py: Webhook handling server
• requirements.txt: Project dependencies

Author

• Email: [[email protected]]
• GitHub: [https://github.com/iamtornado]
• website: [https://alidocs.dingtalk.com/i/nodes/Amq4vjg890AlRbA6Td9ZvlpDJ3kdP0wQ?utm_scene=team_space]

License

This project is licensed under the MIT License. For more information, please see the License file included in this repository.