Nutriscan An AI-Powered Mobile Application for Ingredient-Based Food Quality Analysis

NutriScan is an AI-powered mobile application designed to evaluate the health quality of packaged food products through ingredient analysis. The system combines barcode scanning, OCR-based ingredient extraction, and Large Language Model (LLM)–driven interpretation to provide users with an easy-to-understand health rating and feedback for any food item. The application is developed using Flutter for the mobile frontend and Flask for the backend API, hosted on PythonAnywhere. When a user scans a product’s barcode, the system checks a cloud database to identify whether the product already exists. If not found, NutriScan extracts ingredients from the product label using OCR and sends them to an AI model (Gemini/ChatGPT) for analysis. The AI model evaluates the health impact of each ingredient, generates structured insights, assigns a health rating, and provides warnings for high-risk components. NutriScan stores product data—including ingredients, nutrient facts, barcode, health score, and health recommendations—making future scans faster and improving the system’s accuracy over time. The project aims to empower consumers with transparent, AI-driven nutritional awareness and to demonstrate how LLMs can support public health decisions through intelligent food analysis. This MTech project includes the complete system design, architecture, database model, backend API development, Flutter-based frontend implementation, AI integration pipeline, and evaluation of the ingredient rating methodology. The application showcases a practical and accessible approach to AI-assisted food assessment that can be extended into large-scale health monitoring and food safety initiatives.