README

Food Scanner App

Upload a meal photo and get instant nutrition guidance. A Next.js frontend and a serverless AWS backend combine OpenAI-powered image analysis with account-based scan history.

Next.js TypeScript React Python OpenAI AWS Lambda API Gateway Cognito DynamoDB S3 / CloudFront

What It Does

Upload a meal photo, get a nutrition estimate, and save the result to a personal scan history when signed in. Signed-out visitors can still analyze photos and see useful nutrition guidance - results just stay on-device instead of syncing to an account.

Architecture

Frontend

Next.js static export with React, Cognito auth helpers, and a responsive card-based interface with inline image compression before upload.

Authentication

Amazon Cognito provides sign-in and a signed-out fallback so users can move between private and local-history modes.

Backend

AWS Lambda behind API Gateway, with Cognito JWT verification on protected routes.

Analysis

Lambda sends the image payload to OpenAI for analysis, with fallback heuristic logic when the AI call is unavailable, and returns user-friendly meal guidance.

Storage

DynamoDB stores per-user scan records for signed-in accounts; the browser keeps a local fallback history for people who are not signed in.

Key Endpoints

GET /health: service health check
GET /config: public runtime configuration
POST /analyze: submit a meal photo for analysis
/scans (protected) - load and save account scan history

This README was written from the app's published architecture and stack rather than pulled from a repo file - the source repository does not currently include a README.md.

Open App Story Visual Diagram Back to Portfolio