mumbaihack24

Mumbai hack hackathon backend app (Nodejs, OLAMA, LLMA 3.0)

---

SmartFoodAI Backend 🌿

Empowering Smarter Food Choices with AI

SmartFoodAI is a backend solution designed to process nutritional information from packaged food labels, enabling users to make informed, healthier decisions. This application leverages cutting-edge AI models, including Meta's Llama 3.2, running locally using OLMa for privacy and efficiency.

---

How It Works

1. OCR on Device: The app uses on-device OCR to capture food label information, ensuring user privacy.
2. Backend Processing:
   • API 1 - Health Insights: Evaluates the extracted elements against a trained dataset, providing a binary health suggestion (Healthy or Unhealthy).
   • API 2 - Detailed Descriptions: Analyzes and returns detailed information about the individual components detected in the OCR output.

The backend APIs are optimized for local computation with OLMa, offering a responsive and private experience without reliance on cloud processing.

---

Powered by Meta’s Llama 3.2 🤖

Meta's Llama 3.2 brings advanced AI capabilities that enhance the application:

• Local AI Inference: Ensures speed and privacy with models optimized for on-device use.
• Multimodal Processing: Capable of understanding and analyzing both text and images, crucial for processing food labels.
• Comprehensive Analysis: Extracts meaningful insights, enabling smarter decisions.

For more about Llama 3.2, visit Meta AI.

---

Built for Innovation 🌍

Mumbai Hacks 2024: This project was born at Mumbai Hacks, the largest AI hackathon, recognized by the Guinness World Records. With over 10,000 participants, this event, hosted at ATLAS SkillTech University, showcased cutting-edge AI solutions. SmartFoodAI represents the spirit of this event, demonstrating how AI can address real-world challenges.

---

APIs Overview

1. Health Insights API:

   • Input: OCR data (JSON)
   • Output: Binary health suggestion (Healthy/Unhealthy)

2. Component Details API:

   • Input: OCR data (JSON)
   • Output: Detailed analysis of each component in the product (e.g., nutritional value, potential health impacts).

Technology Stack

• AI Models: Meta Llama 3.2 with OLMa for local execution.
• Backend Framework: Node.js with Express.
• On-Device OCR: Enabled through a lightweight library for maximum privacy.

---

Getting Started

1. Install Dependencies:

   npm install  

2. Run the Server:

   npm start  

3. Use the API endpoints for:
   • Health suggestions.
   • Component-level insights.

---

Contributing

Contributions to improve SmartFoodAI are always welcome! Submit a PR or reach out with your ideas.

---

With SmartFoodAI, we're merging technology and health to create a smarter, healthier future. 🌟