How to Build for the Cloud

1. Start with a web app. You build an app that renders a map of fishing spots with a water temperature overlay. You deploy it to Spawned as a function with a domain connection, reachable at fishmap.spawned.app. Nothing runs between requests, so the app costs nothing while idle. At this point your cloud infrastructure is one function and a hostname.

2. Add user accounts and saved spots. Users want to save their favorite fishing spots. That means you need a database. You add a Database component to your project — from the canvas, the CLI, or by asking your coding agent — and connect it to your function. On the next deploy Spawned provisions a managed RDS instance and injects the credentials into your app. Your app now has compute and a database.

3. Store large map tile files. You want to save historical water temperature data so users can look back over time. Each temperature map is a large collection of tile files, not the kind of data you put in a database. You add a bucket component and connect your function to it, and Spawned provisions the S3 bucket and hands your app its name.

4. Fetch temperature data on a schedule. The temperature data needs to update every hour. That is a scheduled job, not something your web server should handle. You write a small function that fetches the data and add it as a Function component with an hourly schedule. Spawned provisions a scheduled cron job that runs on the hour.

You now have a function serving the web app, a database for user data, a bucket for large files, and a second function running hourly. All declared in one infra.json and provisioned by Spawned. If the web app later needs to hold connections open or run long requests, it moves to a container and the rest stays as it is.