What owning the code actually means

Nearly every software vendor says the client owns the code. It is one of those claims that costs nothing to make and is rarely tested until the relationship ends.

The test is simple: if you fired us tomorrow, could another team deploy this? Answering that honestly usually surfaces three gaps. The repository is ours and you get a copy at the end. The infrastructure sits in our cloud account under our billing. And the deployment process lives in somebody's terminal history rather than in a script anyone can run.

We avoid all three by default. Work happens in your repository from the first commit, so ownership is a fact of the setup rather than a clause. Infrastructure goes in your cloud accounts, billed to you, with us holding access we can lose. Deployment is a script in the repository that a new engineer can run on their first day, because that is the only version of "documented" that survives contact with reality.

None of this is generous. It removes the leverage that makes vendors complacent, which is the point: we would rather be kept because leaving is easy and you did not want to.