AIAssistantWithRAG

A powerful Multi-Company AI Assistant with RAG (Retrieval-Augmented Generation)

Features • Installation • Usage • API • Contributing

Description

AIAssistantWithRAG is a production-ready AI assistant platform designed for multi-tenant environments. It leverages Retrieval-Augmented Generation (RAG) to provide accurate, context-aware responses based on company-specific documents and knowledge bases.

Key Highlights

Multi-Company Architecture: Complete data isolation between companies
RAG-Powered Intelligence: Answers grounded in your actual documents
Multilingual Support: Understands French, English, and Moroccan Darija
Privacy-First: Local LLM with Ollama - your data stays on your servers
Product Catalog Integration: Smart product recommendations with images and pricing
High Performance: Qdrant vector database for lightning-fast semantic search
Docker Ready: Complete containerization with Docker Compose
Cross-Platform: Supports Windows, Linux, and macOS deployment

Architecture

AIAssistantWithRAG/
│
├── main.py                    # FastAPI application entry point
├── config.py                  # Configuration & environment settings
│
├── api/
│   ├── upload.py                 # Document upload & processing endpoints
│   └── chat.py                   # Conversational AI endpoints
│
├── rag/
│   ├── embeddings.py             # BGE-M3 multilingual embeddings
│   ├── qdrant.py                 # Vector database operations
│   └── chain.py                  # LangChain RAG orchestration
│
├── docker/
│   ├── Dockerfile                # Application container
│   ├── docker-compose.yml       # Multi-service orchestration
│   ├── .dockerignore            # Docker build context
│   └── .env.docker              # Docker environment template
│
├── prompts/
│   └── master_prompt.txt         # AI behavior & response templates
│
├── services/
│   ├── document_loader.py        # Multi-format document processing
│   └── product_service.py        # Product search & recommendations
│
├── scripts/
│   ├── start.sh                  # Unix/Linux startup script
│   └── start.bat                 # Windows startup script
│
└── storage/
    └── company_{id}/             # Isolated company document storage

Features

Core Capabilities

Feature	Description
Multi-Tenancy	Complete data isolation per company with unique collections
Document Processing	Support for PDF, TXT, DOCX, CSV, JSON formats
Semantic Search	BGE-M3 embeddings for accurate multilingual retrieval
Conversational AI	Context-aware chat with conversation history
Product Intelligence	Smart product recommendations with pricing
Multilingual	Darija, French, English understanding & responses

Technical Stack

Backend Framework: FastAPI (async, high-performance)
LLM Orchestration: LangChain
Vector Database: Qdrant
Embeddings: BAAI/bge-m3 (1024 dimensions)
Local LLM: Ollama (llama3.2)
Document Processing: LangChain Document Loaders

Prerequisites

Choose your deployment method:

Docker Deployment (Recommended)

Docker and Docker Compose installed
8GB+ RAM recommended
5GB free disk space for models

Manual Installation

Python 3.11+ installed
Docker (for Qdrant) or Qdrant Cloud account
Ollama installed and running
8GB+ RAM recommended for embeddings model

Installation

Option 1: Docker Deployment (Recommended)

Step 1: Clone the Repository

git clone https://github.com/ISTIFANO/AIAssistantWithRAG.git
cd AIAssistantWithRAG

Step 2: Configure Environment

# Copy Docker environment template
cp .env.docker .env

# Edit if needed (optional)
nano .env

Step 3: Start with Docker Compose

# Build and start all services
docker-compose up -d --build

# Or use the startup scripts:
# Windows:
start.bat

# Unix/Linux/macOS:
./start.sh

Step 4: Verify Installation

# Check all containers are running
docker ps

# Check API health
curl http://localhost:8000/health

Your AI Assistant is ready at: http://localhost:8000

Option 2: Manual Installation

Step 1: Clone the Repository

git clone https://github.com/ISTIFANO/AIAssistantWithRAG.git
cd AIAssistantWithRAG

Step 2: Create Virtual Environment

# Create environment
python -m venv venv

# Activate (Windows)
venv\Scripts\activate

# Activate (Linux/Mac)
source venv/bin/activate

Step 3: Install Dependencies

pip install -r requirements.txt

Step 4: Configure Environment

# Copy example configuration
cp .env.docker .env

# Edit with your settings
nano .env  # or use your preferred editor

Step 5: Start Qdrant Vector Database

# Using Docker
docker run -p 6333:6333 -p 6334:6334 qdrant/qdrant

# Or use Qdrant Cloud (update .env accordingly)

Step 6: Setup Ollama LLM

# Install Ollama from https://ollama.ai

# Pull the model
ollama pull llama3.2

# Verify it's running
ollama list

Step 7: Run the Application

# Development mode with auto-reload
uvicorn app.main:app --reload --host 0.0.0.0 --port 8000

# Production mode
uvicorn app.main:app --workers 4 --host 0.0.0.0 --port 8000

Access the API at: http://localhost:8000

API Documentation: http://localhost:8000/docs

Docker Services

The Docker setup includes 4 services:

Service	Description	Port
ai-assistant-app	FastAPI application	8000
ai-assistant-qdrant	Vector database	6333, 6334
ai-assistant-ollama	Local LLM server	11434
ai-assistant-ollama-init	Model downloader	-

Docker Commands

# Start all services
docker-compose up -d

# View logs
docker-compose logs -f

# Stop all services
docker-compose down

# Rebuild and restart
docker-compose up -d --build

# Check service status
docker ps

API Endpoints

Health Check

GET /health

Upload Documents

Upload company documents for RAG processing.

POST /api/v1/upload
Content-Type: multipart/form-data

Parameter	Type	Required	Description
`company_id`	string	Yes	Unique company identifier
`company_name`	string	Yes	Company display name
`files`	File[]	Yes	Documents to upload

Example Response:

{
  "success": true,
  "message": "Successfully processed 3 files",
  "company_id": "cmp_001",
  "files_processed": 3,
  "chunks_created": 127
}

Chat

Conversational endpoint with RAG-powered responses.

POST /api/v1/chat
Content-Type: application/json

Request Body:

{
  "company_id": "cmp_001",
  "company_name": "My Company",
  "message": "شنو العروض اللي عندكم دابا؟",
  "conversation_history": []
}

Response:

{
  "success": true,
  "reply_text": "عندنا عروض مزيانين هاد الشهر! كاين تخفيضات على...",
  "products": [
    {
      "name": "Premium Package",
      "description": "باقة متكاملة للشركات",
      "price": "999 DH",
      "image_url": "https://example.com/image.jpg"
    }
  ],
  "sources": ["offers_2024.pdf", "catalog.pdf"]
}

Delete Company Data

DELETE /api/v1/documents/{company_id}

Configuration

Environment Variables

Variable	Description	Default
`APP_NAME`	Application name	AI Assistant with RAG
`DEBUG`	Debug mode	True
`QDRANT_HOST`	Qdrant server host	localhost
`QDRANT_PORT`	Qdrant server port	6333
`QDRANT_API_KEY`	Qdrant API key (optional)	-
`OLLAMA_BASE_URL`	Ollama API URL	http://localhost:11434
`OLLAMA_MODEL`	LLM model name	llama3.2
`EMBEDDING_MODEL`	Embedding model	BAAI/bge-m3
`EMBEDDING_DIMENSION`	Vector dimensions	1024
`CHUNK_SIZE`	Document chunk size	1000
`CHUNK_OVERLAP`	Chunk overlap	200
`TOP_K_RESULTS`	RAG retrieval count	5

Usage Examples

Python Client

import requests

# Upload documents
files = [
    ('files', open('catalog.pdf', 'rb')),
    ('files', open('pricing.csv', 'rb'))
]
data = {
    'company_id': 'cmp_001',
    'company_name': 'My Company'
}
response = requests.post(
    'http://localhost:8000/api/v1/upload',
    files=files,
    data=data
)

# Chat
response = requests.post(
    'http://localhost:8000/api/v1/chat',
    json={
        'company_id': 'cmp_001',
        'company_name': 'My Company',
        'message': 'What products do you have?'
    }
)
print(response.json())

cURL

# Upload
curl -X POST http://localhost:8000/api/v1/upload \
  -F "company_id=cmp_001" \
  -F "company_name=My Company" \
  -F "files=@catalog.pdf"

# Chat
curl -X POST http://localhost:8000/api/v1/chat \
  -H "Content-Type: application/json" \
  -d '{"company_id":"cmp_001","company_name":"My Company","message":"Hello!"}'

Testing

# Run tests
pytest tests/ -v

# With coverage
pytest tests/ --cov=app --cov-report=html

Supported File Formats

Format	Extension	Description
PDF	`.pdf`	Product catalogs, manuals
Text	`.txt`	Plain text documents
Word	`.docx`	Microsoft Word documents
CSV	`.csv`	Product lists, pricing tables
JSON	`.json`	Structured data, configs

Security Features

Data Isolation: Each company's data stored in separate Qdrant collections
Local Processing: LLM runs locally with Ollama
No Data Leakage: Strict company-based filtering in all queries
Input Validation: Pydantic models for request validation

Contributing

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

Fork the repository
Create your feature branch (git checkout -b feature/AmazingFeature)
Commit your changes (git commit -m 'Add some AmazingFeature')
Push to the branch (git push origin feature/AmazingFeature)
Open a Pull Request

License

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

Author

ISTIFANO

GitHub: @ISTIFANO

Acknowledgments

FastAPI - Modern Python web framework
LangChain - LLM application framework
Qdrant - Vector similarity search engine
Ollama - Local LLM runner
BAAI - BGE embedding models

Star this repo if you find it helpful!

Made with love in Morocco

Name		Name	Last commit message	Last commit date
app		app
.dockerignore		.dockerignore
.env.docker		.env.docker
.env.example		.env.example
.gitignore		.gitignore
Dockerfile		Dockerfile
README.md		README.md
docker-compose.yml		docker-compose.yml
import_company_data.ps1		import_company_data.ps1
import_company_data.py		import_company_data.py
requirements.txt		requirements.txt
start.bat		start.bat
start.sh		start.sh
test_ai_assistant.py		test_ai_assistant.py
test_app.py		test_app.py
test_document.txt		test_document.txt

Latest commit

History

Folders and files

Repository files navigation

AIAssistantWithRAG

Description

Key Highlights

Architecture

Features

Core Capabilities

Technical Stack

Prerequisites

Docker Deployment (Recommended)

Manual Installation

Installation

Option 1: Docker Deployment (Recommended)

Step 1: Clone the Repository

Step 2: Configure Environment

Step 3: Start with Docker Compose

Step 4: Verify Installation

Option 2: Manual Installation

Step 1: Clone the Repository

Step 2: Create Virtual Environment

Step 3: Install Dependencies

Step 4: Configure Environment

Step 5: Start Qdrant Vector Database

Step 6: Setup Ollama LLM

Step 7: Run the Application

Docker Services

Docker Commands

API Endpoints

Health Check

Upload Documents

Chat

Delete Company Data

Configuration

Environment Variables

Usage Examples

Python Client

cURL

Testing

Supported File Formats

Security Features

Contributing

License

Author

Acknowledgments

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages