AI Product Control Agent – Your Modern Copilot

Meet your next-level assistant: Product Control Agent – an AI-powered solution that makes controlling and analyzing your product website smarter and faster.

---

Content Index

• 🚀 Overview
• 🔥 Key Features
• 🛠️ Tech Stack
• 📁 Folder Structure
• ⚡ Getting Started
  • Prerequisites
  • Installation
  • Environment Setup
  • Running the Application
• 🌐 Usage
• 🛡️ Security
• 📊 Monitoring AI Performance
• 🤝 Contributing

---

🚀 Overview

Product Control Agent is built to automate and simplify CRUD operations on your product website, leveraging the latest in AI to streamline workflows and supercharge productivity. With advanced analysis capabilities, it helps you make better decisions, faster. Vist this link to view the application demo.

Note: This repository does not include the Flask backend. Please visit: ai_product_agent to access or refer to the AI Agent implementation.

---

🔥 Key Features

• AI-Driven Automation: Leverages cutting-edge LLMs (like Llama 3 70B) to power intelligent CRUD operations and data analysis.

• Seamless CRUD: Effortlessly manage products, categories, and data through an intuitive interface.

• Rapid Analysis: Instantly generate AI-driven insights from your product data.

• Modern Full-stack Architecture: Built on the robust MERN stack (MongoDB, Express, React, Node.js) with a Flask-based AI backend.

• CORS Support: Enables smooth integration between frontend and backend services.

• JWT Authentication: Ensures secure access to protected API endpoints.

• MongoDB Storage: Scalable and efficient data storage for all your product records.

• LLM-Powered Intelligence: Uses Llama 3 70B (or similar models) for deep data analysis and smart recommendations.

• RESTful API: Provides clean, well-documented endpoints for seamless integration across your frontend and backend.

• Secure Communication: Designed with security best practices for safe data exchange within your stack.

• Configurable AI Model Path: Easily switch or upgrade AI models as needed.

• Flexible Image Generation Options:

  • SerpAPI Integration: Generate images using SerpAPI. Implemented in image_tools.py.
  • Unsplash API Integration: Alternatively, use Unsplash API. Implemented in image_tools2.py.

• Automatic AWS S3 Uploads: Downloaded images are automatically uploaded to your configured AWS S3 bucket. Ensure your AWS credentials are correctly set up.

---

🛠️ Tech Stack

• Frontend: React.js, TailwindCSS (node-jwt-frontend)
• Backend: Node.js, Express, MongoDB, JWT, CORS (nodejs-jwt-auth)
• AI Service: Flask, Llama 3 70B LLM

---

📁 Folder Structure

product_agent_node/
├── node-jwt-frontend/    # React.js + TailwindCSS frontend
├── nodejs-jwt-auth/      # Node.js + Express + MongoDB + JWT + CORS backend
├── flask-ai/             # Flask server for Llama 3 70B (if applicable)
├── README.md
└── ...

---

⚡ Getting Started

Prerequisites

• Node.js (v16+)
• npm or yarn
• MongoDB (local or Atlas)
• Python 3.8+ (for Flask AI service)
• (Optional) CUDA for GPU-accelerated inference

Installation

1. Clone the repository:

   git clone https://github.com/YUGESHKARAN/product_agent_node.git
   cd product_agent_node

2. Install Frontend:

   cd node-jwt-frontend
   npm install
   cd ..

3. Install Backend:

   cd nodejs-jwt-auth
   npm install
   cd ..

4. Install Flask AI Server (optional):

   cd flask-ai
   pip install -r requirements.txt
   cd ..

Environment Setup

• Backend (nodejs-jwt-auth/.env):

  PORT=5000
  MONGODB_URI=mongodb://localhost:27017/your-db
  JWT_SECRET=your_jwt_secret

• Backend (flask-ai/.env):

     GROQ_API_KEY= groq_api_key   # To access the model llama3-70b
     SERPAPI_KEY=your_serpapi_key        # Optional, for SerpAPI image generation
     UNSPLASH_ACCESS_KEY=your_unsplash_key  # Optional, for Unsplash image generation
     AWS_ACCESS_KEY_ID=your_aws_access_key_id      # Required, for S3 image storage
     AWS_SECRET_ACCESS_KEY=your_aws_secret_access_key  # Required, for S3 image storage
     AWS_S3_BUCKET_NAME=your_bucket_name           # Required, for S3 image storage

• Frontend (node-jwt-frontend/.env):

  REACT_APP_API_URL=http://localhost:5000
  

• Flask AI (flask-ai/.env):

  MODEL_PATH=path/to/llama3-70b-8192
  

Running the Application

• Express Backend:

  cd nodejs-jwt-auth
  npm start

• Frontend:

  cd node-jwt-frontend
  npm start

• Flask AI Server:

  cd flask-ai
  python app.py

---

🌐 Usage

• Visit http://localhost:5173 to access the frontend.
• The frontend communicates with the Node.js backend for CRUD operations and with the Flask AI service for advanced analysis.
• All endpoints are secured using JWT tokens.
• The Flask backend exposes endpoints for AI-powered analysis, inference, and image generation.
• Integration between services is done via HTTP requests.

🖼️ Image Generation

• You can choose your preferred image generation backend:
  • SerpAPI: Implemented in image_tools.py – requires a valid SERPAPI_KEY.
  • Unsplash API: Implemented in image_tools2.py – requires a valid UNSPLASH_ACCESS_KEY.
• To switch between backends, update the import in mongodb_database.py:

  from image_tools2 import search_and_download_image

---

🛡️ Security

• The backend is intended for internal use behind your main application’s authentication layer.
• Set up firewall rules and/or JWT authentication if exposing to public networks.

---

📊 Monitoring AI Performance

• Utilizes LangSmith to monitor Product Agent logs, operations, token usage, costs, and responses.

---

🤝 Contributing

Contributions and ideas are welcome! Open issues or submit pull requests.

Pushing the boundaries of what AI can do to simplify workflows and boost productivity!