Six clients, six sets of credentials, and a per-token bill that was impossible to attribute cleanly. The fix wasn't a spreadsheet — it was moving the work onto plans, and keeping every client's access separate.
Agency work makes per-token billing genuinely painful: one invoice, six clients, and no honest way to split it. Worse, the credentials sprawl — a key here, a service account there, and nobody sure which client's stack a given secret can still reach.
Each client's stack gets its own scoped workspace with its own vaulted credentials — nothing pooled and nothing shared between clients. Access is read-only until someone approves a write, and revoking a client's row stops agents using it on the very next run. When an engagement ends, the access ends with it.
Client stacks fail on the things nobody owns: the dangling DNS record, the forgotten staging box, the key that outlived the contractor. Watchers run continuously across all six and surface those the moment they appear, which is a far better story for a client call than discovering it during an audit.
Connect the account you already have, point Cortev at your repos, and let it run while you're somewhere else.
An illustrative scenario of how Cortev is used — a composite, not a named customer, and the figures above are examples rather than audited results. Your own savings depend on your usage and your plan's fair-use limits.