Overview

Foodie AI is a mobile application designed to help users make healthier food choices by leveraging artificial intelligence. By simply taking a photo of a meal, users receive an instant, detailed breakdown of the dish, including ingredients, nutritional values, and potential allergens. The app integrates seamless authentication, real-time data synchronization, and a subscription model for premium insights.

Key Features

• AI-Powered Food Recognition: Utilizes OpenAI's advanced vision models to identify food items from camera captures or gallery uploads with high accuracy.
• Detailed Nutritional Analysis: Provides a comprehensive breakdown of calories, macronutrients, and ingredients, helping users track their daily intake against estimated values.
• Dietary Customization: Allows users to set allergies and health conditions, ensuring the AI highlights potential risks in every scanned dish.
• Smart History & Tracking: Automatically saves scanned dishes to a personal history, enabling users to review their eating habits over time.
• Premium Subscription: Integrated RevenueCat paywall for unlocking advanced features, ensuring a sustainable monetization strategy.
• Secure Authentication: Robust user management powered by Clerk, supporting Email, Google, and Apple sign-ins.

Technologies Used

• React Native (Expo): For building a performant, cross-platform mobile user interface.
• Convex: As the backend-as-a-service (BaaS) for real-time data storage, serverless functions, and seamless state synchronization.
• OpenAI API: Powers the core image analysis and nutritional estimation logic.
• Clerk: Handles secure user authentication and session management.
• RevenueCat: Manages in-app subscriptions and entitlement checks.
• PostHog: Provides product analytics to understand user behavior and improve app flows.
• TypeScript: Ensures code reliability and type safety across the full stack.

Challenges and Learnings

One of the main challenges was orchestrating the asynchronous flow between image upload, external AI processing, and client updates. We implemented a robust system using Convex mutations and actions to handle the "processing" state, giving users immediate feedback while the heavy lifting happened in the background. Additionally, integrating multiple third-party services (Clerk, RevenueCat, OpenAI) required careful handling of API keys and environment environments to ensure security and reliability.

Outcome

Foodie AI has evolved into a polished, functional assistant that bridges the gap between convenience and health awareness. The seamless integration of AI into the daily eating routine offers users immediate value, while the scalable backend architecture ensures the app can grow with its user base. The project demonstrates the power of modern full-stack tools in rapidly deploying complex, AI-driven mobile applications.