High-Performance RAG Application

🚀 Features

📋 Requirements

🏗️ Architecture

┌─────────────────┐    ┌──────────────────┐    ┌─────────────────┐
│   Load Balancer │────│  Auto Scaling    │────│   EFS Storage   │
│      (ALB)      │    │     Group        │    │  (Documents +   │
└─────────────────┘    └──────────────────┘    │   ChromaDB)     │
                                │               └─────────────────┘
                                ▼
                       ┌──────────────────┐
                       │  EC2 Instances   │
                       │  (FastAPI +      │
                       │   LangChain)     │
                       └──────────────────┘

🚀 Quick Start

Local Development

Using Admin Interface

# Interactive mode
python ui/admin_interface.py --interactive

# Command line mode
python ui/admin_interface.py --bulk-setup
python ui/admin_interface.py --add-user user1 pass123
python ui/admin_interface.py --upload-dir test_data shared

🧪 Performance Testing

Realistic Performance Test (Recommended)

# Gradual load test that finds your system's limits
python realistic_performance_test.py

Laptop-Friendly Testing

# Quick test for development laptops
python realistic_performance_test.py --quick

High-Load Testing (Requires Cloud APIs)

# Only works with OpenAI/Gemini API keys
python test_performance.py --users 300 --duration 10 --rpm 1000

Performance Expectations:

☁️ AWS Deployment

Prerequisites

# Install AWS CLI and configure
aws configure
aws sts get-caller-identity  # Verify setup

Automated Setup

cd deploy
python aws_config.py        # Generate AWS configuration
python docker_build.py      # Build and push Docker image

Manual AWS Setup

📊 API Endpoints

Authentication

User Management

File Management

Query

Example Usage

import requests

# Add user
response = requests.post(
    "http://localhost:8000/admin/users/add",
    auth=("admin", "admin"),
    data={"user_id": "newuser", "password": "newpass"}
)

# Upload file
with open("document.pdf", "rb") as f:
    response = requests.post(
        "http://localhost:8000/admin/files/upload",
        auth=("admin", "admin"),
        files={"file": f},
        data={"user_id_for_file": "newuser"}
    )

# Query
response = requests.post(
    "http://localhost:8000/query/",
    auth=("newuser", "newpass"),
    data={"query": "What is this document about?"}
)

🔧 Configuration

Environment Variables

# AI API Keys
OPENAI_API_KEY="your-openai-key"
GOOGLE_API_KEY="your-gemini-key"

# Model Configuration
OPENAI_MODEL="gpt-4o"
GEMINI_MODEL="gemini-pro"
OLLAMA_MODEL="gemma:2b"

# AWS Configuration
AWS_REGION="ap-south-1"
EFS_FILE_SYSTEM_ID="fs-xxxxxxxxx"

Rate Limiting

@limiter.limit("1500/minute")  # Adjust as needed

Scaling Configuration

📁 Project Structure

rag_project/
├── backend/                 # FastAPI application
│   ├── main.py             # Main application
│   ├── services.py         # Core services
│   ├── requirements.txt    # Python dependencies
│   ├── Dockerfile          # Container configuration
│   └── .env               # Environment variables
├── client/                 # Client tools
│   └── client.py          # Command-line client
├── ui/                     # User interfaces
│   └── admin_interface.py # Admin interface
├── deploy/                 # Deployment tools
│   ├── aws_config.py      # AWS configuration helper
│   ├── docker_build.py    # Docker build script
│   └── autoscaling_template.json
├── test_data/             # Test documents and questions
├── test_performance.py    # Performance testing tool
├── DEPLOYMENT_GUIDE.md    # Detailed deployment guide
└── README.md             # This file

🔍 Monitoring and Troubleshooting

Health Check

curl http://your-server:8000/docs

Logs

# Docker logs
docker logs rag-app

# Performance test logs
tail -f performance_test.jsonl

Common Issues

💰 Cost Estimation

10-minute test (300 users, 1000 RPM):

Monthly production (moderate load):

🤝 Contributing

📄 License

🆘 Support

🔮 Roadmap