AgentFlow

An intelligent multi-agent Django application powered by LangGraph and Google Gemini, featuring role-based access control and specialized AI agents for document management and movie discovery.

🌟 Features

• Multi-Agent Supervisor System: Intelligent routing between specialized agents using LangGraph
• Document Management Agent: Full CRUD operations with permission-based access control
• Movie Discovery Agent: Integrated TMDB API for movie search and detailed information
• Role-Based Access Control (RBAC): Powered by Permit.io for fine-grained permissions
• AI-Powered Intelligence: Google Gemini LLM for natural language understanding
• Interactive Development: Jupyter notebooks for experimentation and testing

🏗️ Architecture

User Request
       ↓
Supervisor Agent (Router)
       ↓
    ┌──┴──┐
    ↓     ↓
 Document  Movie
 Agent     Agent

The supervisor agent intelligently routes user requests to the appropriate specialized agent based on the query context.

🚀 Quick Start

Prerequisites

• Python 3.12+
• Django 5.0+
• Google Gemini API Key
• TMDB API Key
• Permit.io API Key

Installation

1. Clone the repository

git clone <your-repo-url>
cd DJANGO-AI-AGENT

2. Create and activate virtual environment

python -m venv venv_aiagent
source venv_aiagent/bin/activate  # Linux/Mac
# or
venv_aiagent\Scripts\activate  # Windows

3. Install dependencies

pip install -r requirements.txt

4. Set up environment variables

Create a .env file in the src directory:

# Django
SECRET_KEY=your-django-secret-key
DEBUG=True

# Google Gemini
GOOGLE_API_KEY=your-gemini-api-key

# TMDB API
TMDB_API_KEY=your-tmdb-api-key

# Permit.io
PERMIT_API_KEY=your-permit-api-key
PERMIT_PDP_URL=https://cloudpdp.api.permit.io

5. Run migrations

cd src
python manage.py migrate

6. Create a superuser

python manage.py createsuperuser

7. Run the development server

python manage.py runserver

📁 Project Structure

DJANGO-AI-AGENT/
├── src/
│   ├── AI/
│   │   ├── agents.py           # Agent definitions
│   │   ├── supervisors.py      # Supervisor routing logic
│   │   ├── llms.py            # LLM configurations
│   │   └── tools/
│   │       ├── documents.py    # Document management tools
│   │       └── movie_discovery.py  # Movie discovery tools
│   ├── documents/             # Document app
│   ├── cfehome/              # Django project settings
│   ├── mypermit/             # Permit.io client
│   └── tmdb/                 # TMDB API client
├── notebook/                 # Jupyter notebooks for testing
└── requirements.txt

🤖 Agents

Document Agent

Manages document operations with permission checks:

• Search documents: Query-based document search
• Create document: Add new documents
• Read document: Retrieve document details
• Update document: Modify existing documents
• Delete document: Remove documents

Movie Discovery Agent

Provides movie information via TMDB API:

• Search movies: Find movies by title/keyword
• Movie details: Get comprehensive movie information including ratings, cast, and synopsis

🔐 Permissions & Roles

The system uses Permit.io for RBAC with the following default roles:

Manager Role

• Full document access (read, create, update, delete)
• Movie discovery access (search, detail)

Viewer Role

• Read-only document access
• Movie discovery access (search, detail)

Configure roles and permissions in the Permit.io dashboard or via the API.

📓 Jupyter Notebooks

Explore the notebook/ directory for interactive examples:

• 1-django-users-perms.ipynb - User and permission basics
• 3-langgraph-django-tools.ipynb - LangGraph integration
• 5-ai-agent.ipynb - Agent implementation
• 6-memory-agent.ipynb - Agent memory management
• 7-agent-crud.ipynb - CRUD operations with agents
• 9-movie-discovery-ai-agent.ipynb - Movie agent examples
• 10-multiagent.ipynb - Multi-agent system
• 11-roles-and-permissions.ipynb - RBAC configuration

🛠️ Usage Examples

Using the Supervisor Agent

from AI.supervisors import get_supervisor
from langchain_core.messages import HumanMessage

supervisor = get_supervisor()

# Document query
response = supervisor.invoke({
    "messages": [HumanMessage(content="Show me my recent documents")]
}, config={"configurable": {"user_id": "user_123"}})

# Movie query
response = supervisor.invoke({
    "messages": [HumanMessage(content="Find movies about space exploration")]
}, config={"configurable": {"user_id": "user_123"}})

Direct Agent Usage

from AI.agents import get_document_agent, get_movie_discovery_agent

# Document agent
doc_agent = get_document_agent()
result = doc_agent.invoke({
    "messages": [HumanMessage(content="List my documents")]
}, config={"configurable": {"user_id": "user_123"}})

# Movie agent
movie_agent = get_movie_discovery_agent()
result = movie_agent.invoke({
    "messages": [HumanMessage(content="Search for Inception")]
}, config={"configurable": {"user_id": "user_123"}})

🔧 Configuration

LLM Model

Configure the Gemini model in AI/llms.py:

def get_gemini_model(model=None):
    return ChatGoogleGenerativeAI(
        model=model or "gemini-2.0-flash-exp",
        temperature=0.7
    )

Supervisor Routing

Customize routing logic in AI/supervisors.py by modifying the ROUTER_SYS message and decision logic.

📦 Dependencies

• Django 5.0+ - Web framework
• LangGraph - Agent orchestration
• LangChain - LLM framework
• Google Generative AI - LLM provider
• Permit.io - Authorization as a service
• TMDB API - Movie database
• Jupyter - Interactive development

🧪 Testing

Run Django tests:

python manage.py test

Use Jupyter notebooks for interactive testing and experimentation.

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📄 License

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

🙏 Acknowledgments

• LangGraph for agent orchestration
• Google Gemini for AI capabilities
• Permit.io for authorization management
• TMDB for movie data

Built with ❤️ using Django, LangGraph, and AI