🚀 Multi-Model AI Microservices Project

🌟 Overview

Welcome to the Multi-Model AI Microservices Project! This project integrates three independent AI-powered services that can be deployed together seamlessly:

• Recommender Service – Suggests products based on user history.
• Fraud Detection Service – Detects suspicious transactions.
• Chatbot Service – Engages users in conversation.

Each service is built using FastAPI and is fully containerized with Docker for easy deployment. You can run everything locally using Docker Compose or deploy it to the cloud with Kubernetes.

---

🔧 Setup & Installation

1️⃣ Install Dependencies

Ensure Python is installed, then run:

pip install -r requirements.txt

2️⃣ Download Pre-trained Models

Run the following script to fetch the necessary AI models:

python scripts/download_models.py

This ensures that each service has the correct model files before running.

3️⃣ Start Services with Docker Compose

Launch all three services at once:

docker compose up --build

This will spin up the services on the following ports:

• Recommender Service → http://localhost:8001
• Fraud Detection Service → http://localhost:8002
• Chatbot Service → http://localhost:8003

---

🔍 Testing the APIs

You can test each service using cURL or Postman.

Test Recommender Service

curl "http://localhost:8001/recommend?user_id=123"

Test Fraud Detection Service

curl -X POST "http://localhost:8002/detect_fraud" \
     -H "Content-Type: application/json" \
     -d '{"amount": 1000, "location": "NY"}'

Test Chatbot Service

curl -X POST "http://localhost:8003/chat" \
     -H "Content-Type: application/json" \
     -d '{"user_input": "Hello!"}'

---

☁️ Deploying with Kubernetes

For cloud deployment, apply the Kubernetes configuration:

kubectl apply -f k8s/

---

Hardware Requirements

To run AI-Orchestrator efficiently, ensure your system meets the following hardware specifications:

1. Processor (CPU/GPU)

• CPU: Multi-core processor (4+ cores, e.g., Intel i7/AMD Ryzen 7) for handling concurrent requests.
• GPU (Optional): Recommended for AI models requiring acceleration (e.g., deep learning). An NVIDIA GPU with CUDA support (8GB+ VRAM) is ideal.

2. Memory (RAM)

• Minimum: 16GB RAM
• Recommended: 32GB RAM for handling multiple AI services efficiently.

3. Storage

• Minimum: 50GB free space per model
• Recommended: SSD (NVMe preferred) for faster data access and improved system responsiveness.

4. Networking

• Stable high-speed internet (especially if using cloud-based AI services or real-time processing).

5. Operating System

• Linux (Ubuntu 20.04+ recommended) or Windows 10/11 with WSL2 for better compatibility with Docker and Kubernetes.

6. Software Dependencies

Ensure the following are installed:

• Docker (Latest version) for containerized deployments.
• Kubernetes (Optional, for orchestration at scale).
• Python 3.8+ for AI model execution and service scripts.
• GPU Drivers & CUDA (If using GPU for acceleration).

7. Scalability Considerations

For production or large-scale deployments:

• Consider cloud-based solutions (AWS/GCP/Azure) for dynamic scaling.
• Use load balancing and Kubernetes clusters to optimize resource utilization.

---

⚠️ Important Notes

• Ensure Docker Desktop is running before using docker compose up.
• Before downloading models, update scripts/download_models.py with the correct model URLs.
• You can modify the Dockerfile to match your specific requirements.

---