Multi-Agent AI Assistant

A production-ready multi-agent AI system powered by AutoGen and GPT-4o, featuring specialized autonomous agents for weather, search, email, and calendar management with an interactive CLI interface.

Preview

Interactive terminal interface built with Rich library, featuring conversation history, session management, and beautiful formatting

Features

Custom Enhancements (My Contributions)

This project builds upon AutoGen fundamentals with significant custom implementations:

• Interactive CLI Interface - Interactive terminal-based chat with Rich library (full custom implementation)
• Conversation Memory System - SQLite-backed persistent conversation history with session management
• Modern Configuration Management - Environment-based config with validation and dotenv support
• Structured Logging - Comprehensive logging system with file + console outputs
• Modular Architecture - Clean separation of concerns with organized package structure
• Robust Error Handling - Graceful error handling and user-friendly error messages
• Comprehensive Documentation - Detailed inline documentation and developer guides

Multi-Agent System

• WeatherAssistant - Real-time weather data, forecasts, and precipitation information
• SearchAssistant - Web search and deep research using Tavily API
• EmailAssistant - Gmail integration for reading, searching, drafting, and sending emails (coming soon)
• CalendarAssistant - Google Calendar management for event scheduling (coming soon)

Technical Stack

• AI Framework: AutoGen 0.6.1 (agentchat, core, ext)
• LLM: OpenAI GPT-4o
• APIs: Tavily Search, Open-Meteo Weather, Gmail, Google Calendar
• Backend: FastAPI 0.115
• CLI: Rich 13.9.4
• Language: Python 3.12

Architecture

4-layer architecture: User Interface → Memory & Persistence → Orchestration → Agents → External Services

The system follows a clean layered architecture:

• User Interface Layer: Interactive CLI (Rich library) and future Slack Bot integration
• Memory & Persistence Layer: SQLite-backed conversation storage with session management
• Orchestration Layer: Multi-agent coordinator using AutoGen's MagenticOneGroupChat with GPT-4o
• Agent Layer: Specialized agents (Weather, Search, Email) with domain-specific tools
• External Services Layer: API integrations (Open-Meteo, Tavily, Gmail, OpenAI)

Quick Start

Prerequisites

• Python 3.12+
• OpenAI API Key
• Tavily API Key
• (Optional) Google Cloud credentials for Gmail/Calendar

Installation

1. Clone the repository

git clone https://github.com/Ctt011/autogen-multiagent-ai-assistant.git
cd autogen-multiagent-ai-assistant

2. Navigate to code directory

cd code

3. Create virtual environment

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

4. Install dependencies

pip install -r requirements.txt

5. Configure environment

# Copy example environment file
cp .env.example .env

# Edit .env and add your API keys
# Required:
#   - OPENAI_API_KEY
#   - TAVILY_SEARCH_KEY
# Optional:
#   - SLACK_BOT_TOKEN (for Slack integration)
#   - GOOGLE_CREDENTIALS_FILE (for Gmail/Calendar)

6. Run the assistant

python main.py

Usage

Interactive CLI

The CLI provides a beautiful terminal interface for chatting with the assistant:

Multi-Agent AI Assistant

Welcome! I'm your AI assistant powered by multiple specialized agents...

Commands:
  /help   - Show help message
  /agents - List available agents
  /clear  - Clear screen
  /quit   - Exit assistant

You> What's the weather in Paris?

Assistant
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Current weather in Paris:
• Temperature: 15°C
• Conditions: Partly cloudy
• Wind: 12 km/h
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

You> Latest news on AI developments

Assistant
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Answer: Recent developments in AI include...

Sources:
1. OpenAI announces GPT-5
   https://example.com/...
   Summary: OpenAI has revealed details...
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Available Commands

Command	Description
/help	Display help message with examples
/agents	List all available agents and their capabilities
/history	View conversation history for the current session
/stats	Display conversation memory statistics
/clear	Clear the terminal screen
/quit or /exit	Exit the assistant

The /agents command shows available AI assistants, and /stats displays conversation memory analytics

Example Queries

Weather Information:

• "What's the weather in London?"
• "Give me a 3-day forecast for Mumbai"
• "Will it rain tomorrow in Tokyo?"
• "Current temperature in New York"

Web Search & Research:

• "Latest developments in artificial intelligence"
• "Research quantum computing applications"
• "What's happening in tech news today?"
• "Find information about AutoGen framework"

Project Structure

autogen-multiagent-ai-assistant/
├── code/                          # Main application code
│   ├── src/                       # Source package
│   │   ├── __init__.py
│   │   ├── config.py              # Configuration management
│   │   ├── logger.py              # Logging setup
│   │   ├── agents.py              # Multi-agent orchestration
│   │   ├── cli.py                 # Interactive CLI interface
│   │   └── tools/                 # Agent tools
│   │       ├── __init__.py
│   │       ├── weather.py         # Weather API integration
│   │       └── search.py          # Tavily search integration
│   ├── main.py                    # Application entry point
│   ├── requirements.txt           # Python dependencies
│   ├── .env.example              # Example configuration
│   └── README_DEV.md             # Developer documentation
├── .gitignore                     # Git ignore patterns
└── README.md                      # This file

Configuration

All configuration is managed through environment variables in the .env file:

Variable	Required	Description
OPENAI_API_KEY	Yes	OpenAI API key for GPT-4o
TAVILY_SEARCH_KEY	Yes	Tavily API key for web search
OPENAI_MODEL	No	Model to use (default: gpt-4o)
OPENAI_TEMPERATURE	No	Temperature setting (default: 1.0)
DEFAULT_TIMEZONE	No	Default timezone (default: America/Los_Angeles)
LOG_LEVEL	No	Logging level (default: INFO)
SLACK_BOT_TOKEN	No	Slack bot token (optional)
GOOGLE_CREDENTIALS_FILE	No	Google API credentials (optional)

Development

Running Tests

# Install development dependencies
pip install pytest pytest-asyncio pytest-cov

# Run tests
pytest tests/

Code Quality

# Format code
black src/

# Type checking
mypy src/

# Linting
pylint src/

Roadmap

Phase 1: Core Features (Current)

• Interactive CLI interface
• Weather assistant
• Search assistant
• Configuration management
• Logging system

Phase 2: Extended Features

• Email assistant (Gmail integration)
• Calendar assistant (Google Calendar)
• Conversation memory/history
• Multi-LLM support (Claude, Gemini)

Phase 3: Advanced Features

• Slack bot interface
• Web UI with FastAPI
• Voice interface integration
• Agent performance analytics
• Comprehensive test suite
• CI/CD pipeline

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• AutoGen - Multi-agent framework
• OpenAI - GPT-4o language model
• Tavily - Search API
• Open-Meteo - Free weather API
• Rich - Beautiful terminal formatting

Contact

Camille Tran - @Ctt011

Project Link: https://github.com/Ctt011/autogen-multiagent-ai-assistant

---

Built using AutoGen and GPT-4o