Skip to main content

Use cases: start where you are

You made something. Infrastream makes it real — built from parts that arrive already wired together and locked down, with the price shown before you say yes, and an undo that is the same button as undoing code.

That sentence is the whole product. The eighteen journeys below are that sentence, growing up.


Pick your entry point​

🚀
Stories 1–9
I built an app and need it live
Solo builders & vibecoders

You already have the app. This arc is the other half: getting it online, on your own domain, with real logins and a real database — without learning a cloud, and without an invoice that surprises you.

Start with Sam's first deploy →
👥
Stories 10–14
We're a team now and I'm the bottleneck
Growing engineering teams

Every small infrastructure request routes through one person. This arc removes that person from the critical path: the agent drafts the change, anyone on the team reviews it, and staging stops being a reconstruction from memory.

Stop being the human API →
🏛️
Stories 15–18
We have an estate and it needs governing
Enterprise & regulated industries

A decade of infrastructure nobody wants to touch, and an access spreadsheet that is wrong more often than right. This arc brings it under management one project at a time — nothing moves, nothing has a scheduled outage.

Governance without a cutover →

How to read these​

Every story is written twice, and you are meant to read only the half that is yours.

In plain words describes the situation and the outcome with no jargon. If you are deciding whether Infrastream fits — as a founder, a product owner, or someone who has never opened a cloud console — these blocks are the whole document. You can skip everything else.

What you actually write and What the platform handles for you are the engineering layers: the real manifest, the real field names, the real resources provisioned. Nothing has been softened to make the plain-language half work.

Each story also carries a short table of what it needs, how long it takes, and what it costs. Cost is never a dollar figure here, because it depends on your region and your usage. It is a shape — "nothing extra", or "one small always-on component", or "worth putting on an off-hours schedule".

The cheapest infrastructure is the kind that is switched off

Several stories provision resources that bill by the hour. Non-production copies of them do not need to exist overnight or at weekends. See Controlling Costs for how a project puts itself to sleep on a schedule and wakes up before you do.


The through-line​

The stories build on one another — Sam's landing page becomes Sam's app becomes Sam's team's platform — but each one stands alone. Read the one closest to where you are today.

Whichever you pick, the same three guarantees hold:

The price is on the change before you approve it. Not on a bill four weeks later. If a change makes something more expensive, that shows up as a number in the thing you are reviewing.

Undo is the same button as undoing code. Infrastructure is a commit. Reverting the commit reverts the infrastructure. There is no separate recovery procedure to remember under pressure.

Nothing runs with more access than the line that granted it. Including the AI agent. Its power is exactly what was written down for it and reviewed — not a broad role it was trusted not to overstep.


Where to go next​