Building an image processing app with GraphQL and async serverless

This post discusses how to build an image processing app using GraphQL and serverless functions. The app allows users to upload images and convert them to sepia tone asynchronously. It uses Hasura GraphQL Engine, a real-time GraphQL server over Postgres, and an event sourcing system to trigger external webhooks on mutations. The backend architecture follows the 3factor.app pattern, which includes real-time GraphQL, reliable eventing, and async serverless. The frontend has two screens: Upload and Convert. The upload screen inserts the image URL into the database with a GraphQL mutation and redirects to the convert screen with the image ID as a parameter. The convert screen waits for the processed image using GraphQL subscriptions and renders it when available.