Cloud Foundations: Where Your API Actually Runs

You've designed a lovely API. Now — where does it run? It needs a computer that's always on, connected to the internet, that can grow when lots of people show up. That's what "the cloud" provides. This course demystifies it, so a beginner can confidently pick where to build without getting lost in three providers' worth of jargon.

Strip away the mystique and the cloud is simple: instead of buying computers, running them in a room, and maintaining them yourself, you rent them from a company that runs enormous data centres — and pay only for what you use. Need a server for an hour? Rent it for an hour. Suddenly have a million users? Rent more, instantly. Traffic drops? Give them back.

This is transformative for a builder. You don't need money for hardware, a server room, or a team to keep machines alive. A person with a laptop can launch something the whole world can reach, and only pay as it grows. The cloud is what makes "anyone can build" actually true.

Cloud comes in layers, and the only real question each one answers is: how much do you want to handle yourself, versus let the provider handle? Less control means more convenience — and usually a faster path to shipping.

• IaaS (Infrastructure as a Service) — you rent a bare machine and manage almost everything on it: the operating system, updates, your app. Most control, most work. Example: a virtual server (AWS EC2, Azure VM).
• PaaS (Platform as a Service) — you just bring your code; the provider handles the machine, the operating system, scaling, and patching. Far less work, a little less control. Example: deploy code to Vercel, Google App Engine, Azure App Service.
• SaaS (Software as a Service) — you don't build anything; you just use finished software. Example: Gmail, Notion, Slack. (This is the layer your own product might one day be, for your users.)

There's an option that sits even higher than PaaS for many tasks, with a slightly confusing name. Serverless doesn't mean there are no servers (there are). It means you never think about them. You hand the provider a single function — "when a request comes in, run this" — and it takes care of everything: starting machines, scaling to millions, and shutting down when idle. You pay only for the moments your code actually runs.

For a beginner building an API, this is often the gentlest possible start. No server to set up, no machine to keep alive, automatic scaling, and near-zero cost when nobody's using it. Many modern APIs are simply a set of these functions.

The "big three" cloud providers — Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP) — can feel overwhelming, each with hundreds of services and their own names for everything. Here's the secret that calms it down: they offer the same core building blocks, just with different labels.

Learn the concepts — a server, a function, file storage, a database — and you can move between providers; the names change, the ideas don't.