Multi-Agent Chatbot System

A sophisticated multi-agent chatbot system that leverages Ollama for intelligent conversations across multiple domains. The system includes specialized agents for general queries, Concordia University Computer Science admissions, and AI-related questions.

Features

• Multi-agent Architecture: Different agents handle specific domains
  • General Knowledge Agent
  • Concordia CS Admission Agent
  • AI Specialist Agent
• Context-aware Conversations: Maintains conversation history and context
• Knowledge Integration: Uses Wikipedia and domain-specific knowledge bases
• Multi-turn Conversations: Supports ongoing dialogues with memory

Prerequisites

• Python 3.8+
• Ollama installed and running locally (default port: 11434)
• The Mistral model loaded in Ollama

Installation

1. Clone the repository:

git clone <repository-url>
cd multiagent_chatbot

2. Install dependencies:

pip install -r requirements.txt

3. Make sure Ollama is running with the Mistral model:

ollama run mistral

Usage

1. Start the FastAPI server:

python main.py

2. The API will be available at http://localhost:8000

API Endpoints

• POST /chat

  • Send a message to the chatbot
  • Request body:

    {
        "text": "your message here",
        "context": {} // optional
    }

• GET /conversation_history/{agent_name}

  • Retrieve conversation history for a specific agent
  • agent_name: "general", "admission", or "ai"

• POST /clear_history/{agent_name}

  • Clear conversation history for a specific agent
  • agent_name: "general", "admission", or "ai"

Example Usage

import requests

# Send a message
response = requests.post(
    "http://localhost:8000/chat",
    json={"text": "What are the admission requirements for Concordia CS?"}
)
print(response.json())

# Get conversation history
history = requests.get("http://localhost:8000/conversation_history/admission")
print(history.json())

Architecture

The system consists of:

1. Base Agent: Abstract base class with common functionality
2. Specialized Agents:
   • General Knowledge Agent: Handles general queries using Wikipedia
   • Admission Agent: Specializes in Concordia CS admission information
   • AI Agent: Handles AI-related queries with technical expertise
3. FastAPI Application: Coordinates between agents and provides REST API

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

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