Getting started
Your first real AWS deploy and teardown with Korve
Go from an empty canvas to running AWS infrastructure and back to nothing — a safe, five-minute first deploy using dry runs so you never create something by surprise.
The scary part of any infrastructure tool is the first time you hit "deploy" against a real account. Korve is built so that moment is boring: you always see exactly what will happen first, and tearing down is as direct as building. Here's the whole loop.
Use a throwaway account for your first run
Before anything, a habit worth keeping: do your first real deploy on a fresh AWS account with a small budget alarm set, not your production one. Everything below works the same, and you sleep better. Attach a scoped IAM policy (see least privilege) and a $1–5 budget.
1. Build a tiny stack
Connect AWS, then drag two or three nodes onto the canvas — say an S3 bucket and an EC2 instance — and draw a connection between them. Configure the key fields. This is a document; nothing exists in AWS yet.
2. Dry run first — always
Open Deploy and run the dry run. This changes nothing in AWS. It returns a per-node ledger: what will be created, what already exists, what will be validated, and — importantly — the IAM roles and policies each connection will generate. Read it. This is your chance to catch a surprise before it's real.
3. Deploy for real
Happy with the plan? Deploy. Korve provisions each node, then wires the connections (creating the scoped IAM). The nodes show live sync status as they come up. Open the AWS console in another tab if you want to see them appear there too — they're real resources, not a simulation.
For services that need a database password (RDS, Aurora, and friends), Korve generates a strong one and shows it to you once in the deploy ledger. Save it then; it isn't stored.
4. Confirm, then tear down
Once you've seen it work, tearing down is the same motion in reverse. Delete the nodes and redeploy, or terminate resources from the resource list. Because each connection remembers the IAM it created, teardown unwinds those permissions too instead of leaving orphaned roles.
If you tagged your test resources (Korve tags what it creates), you can also sweep everything by tag afterward to be certain nothing lingers billing you.
The takeaway
The dry-run-then-deploy loop means you never create infrastructure by accident, and the delete-and-redeploy loop means you never leave it running by accident. That's the whole safety story, and it's why a first deploy here isn't a leap of faith.
Next
Build this on a canvas instead of the console.
Start Korve free →