This is a full-stack, microservices-based application designed to provide intelligent compliance consulting services. It features an AI-powered chat interface that helps users navigate regulatory requirements by leveraging a comprehensive knowledge base of regulations and laws.

• Architecture
• Features
• Technology Stack
• Project Structure
• Getting Started
• Configuration
• Development
• License

The system is built with a microservices architecture, containerized with Docker, and orchestrated with Docker Compose.

• Frontend: A Next.js application serving the user interface. It provides a modern, responsive dashboard for users to interact with the AI assistant.
• Backend: A Django REST API that manages business logic, user authentication, chat sessions, and communication with the AI service.
• AI Service (LightRAG): A dedicated service that integrates with knowledge graphs (Neo4j), vector databases (ChromaDB), and Large Language Models (LLMs) to provide intelligent responses.
• Databases:
  • PostgreSQL: The primary relational database for the backend service.
  • Neo4j: A graph database for storing and querying complex regulatory relationships.
  • ChromaDB: A vector database for efficient similarity searches required for embeddings.
• Reverse Proxy (Nginx): An Nginx server that acts as a reverse proxy, routing incoming traffic to the appropriate frontend or backend service.

• AI-Powered Chat Interface: An interactive chat for compliance-related queries.
• Comprehensive Knowledge Base: Manages a vast collection of regulations and laws.
• User Authentication: Secure registration and login system.
• Thread-Based Conversations: Organizes user interactions into distinct conversation threads.
• Multi-language Support: Fully localized for English and Persian (Farsi).
• Modern UI/UX: A responsive and intuitive dashboard built with modern web technologies.
• RESTful API: A well-structured API for seamless frontend-backend communication.

The monorepo is organized into the following key directories:

.
├── backend/         # Django REST API service
├── frontend/        # Next.js web application
├── deployment/      # Docker Compose and environment configurations
├── nginx/           # Nginx reverse proxy configuration
├── LightRAG/        # AI service submodule/code (RAG with graphs)
├── data/            # Raw scraped & test data
├── databases/       # Local data persistence for databases (mounted via volumes)
└── LICENSE          # Project license

• Docker
• Docker Compose

1. Clone the repository:

   git clone https://github.com/your-username/compliance-ai-assistant-system.git
   cd compliance-ai-assistant-system

2. Configure Environment Variables: The system relies on environment variables for configuration. The main configuration file for the AI service is deployment/lightrag.env. You may need to provide your own API keys for LLM and embedding services.

3. Run the application: Use Docker Compose to build and run all the services.

   cd deployment
   docker-compose up --build -d

4. Access the application:

   • The frontend application will be available at http://localhost:8000.
   • The backend API is accessible at http://localhost:8000/api/.
   • The LightRAG service runs on port 9621.

• Service Configuration: The deployment/docker-compose.yaml file defines all the services, networks, and volumes.
• AI Service: The deployment/lightrag.env file contains critical settings for the LightRAG service, including API keys for LLMs, database connections, and model parameters.
• Nginx: The reverse proxy is configured in nginx/default.conf.

For detailed instructions on how to set up a local development environment for the frontend and backend, please refer to their respective README files:

• Backend README
• Frontend README

This project is licensed under the MIT License.