In production today
What keeps your record honest.
Tamper-evident audit trail
Every action, by a person or the AI, goes into a log that can only be added to. Any tampering shows. You can prove who did what, and when.
Digital signatures
Permits and approvals are signed on screen, and every signature is tied to that same audit trail.
Access that knows who’s qualified
It knows each person’s role, qualifications and how rested they are. Someone unqualified can’t be put on the job.
Two-factor sign-in
Every account needs an authenticator-app code to sign in.
Works without a signal
Work orders and permits keep working on the device with no link, and catch up when the signal returns.
The autonomy gate
You choose, task by task, whether the AI may suggest, prepare or act. Out of the box, it only suggests.
Certification
In testing now. Certification is our 2027 programme.
The straight answer: we do not yet hold SOC 2 or ISO 27001 certification, and we do not claim API 580 alignment as a certified standard. The controls behind all three are in testing now. Formal certification is our 2027 programme. Until then, we will sit down with your security team and show them how it is built and how the audit record works.
- SOC 22027
- ISO 270012027
- API 580 alignment2027
How AI is kept in check
Nothing acts until you say so.
For each task, you decide whether AI may suggest, prepare or act. It earns more say only by being measurably right, and loses it automatically when its accuracy slips. Safety-critical calls always stay with a person. Everything AI does lands in the same tamper-evident record as your crew’s work.
What we won’t do
We will not train foundation models on your data. Ever.
We will not ship an AI feature without written rules and a way to undo it.
We will not fake demos. Every screen you see is drawn from the real product.
We will not claim a capability without its tag — LIVE, TESTING, BUILD, SPEC or OPEN.