Pilot check

Will Your AI Pilot
Reach Production?

Ten questions, two minutes, scored in your browser. Nothing is sent anywhere. The questions are the controls we run our own company on, and the ones we see missing from pilots that stalled.

01 Is there a written list of what the system may do on its own and what has to wait for a person?
02 When it wants to do something outside that list, does it stop and ask in one place somebody checks?
03 Is there a daily spend limit it cannot pass, enforced in code rather than by someone watching a dashboard?
04 Can one person halt every automated run with one switch, without calling an engineer?
05 Does it read from and write to the tools your team already uses, rather than a copy of the data?
06 Is every action it takes recorded with its input and its result, in a form someone could read next week?
07 Do you know what it costs per finished piece of work, as opposed to per month?
08 Has it run for a full week with nobody watching, and did someone check the output afterwards?
09 Is there a named owner whose job includes reviewing what it did and switching off what stopped helping?
10 When it gets something wrong, is the mistake reversible, or caught before it reaches a customer or a bank account?
0 of 10 answered

What the Score Measures

A pilot dies at the first question a responsible person asks. What happens when it is wrong? Who checks it? What does it cost if it runs all weekend, and how do I stop it? Each question above is one of those, put to your system.

The five areas are the approval line, cost control, the stop switch with its audit trail, integration with the systems you already run, and the plain matter of whether anyone owns it. We score our own operations against the same list, in public, on the how we run page.