A1
No unreviewed worker evaluation
Nothing that ranks or penalises a named person reaches anyone without a human in the path.
Eight rules checked where the record is written. Not a policy PDF. Not a training slide. A gate the log will not cross.
[01 / WHY]
REVEAL reads conversations that affect people’s jobs, customers’ money and regulators’ trust. The constitution is the short list of things the platform is not allowed to do, even when a client asks and even when the model would oblige.
Each rule is checked at write time. An output that breaks one is refused. There is no quiet flag after the fact, and no override that leaves the broken record in the chain.
[02 / THE EIGHT]
The same list the audit chain page quotes. Here it is the whole page, because the rules are the product’s spine.
A1
Nothing that ranks or penalises a named person reaches anyone without a human in the path.
A2
Below the confidence threshold, the result says so and asks for review rather than presenting itself as settled.
A3
Numbers come from a computation, quotes come from the transcript. Where there is nothing, the answer is nothing.
A4
Personal data is redacted by default, on the way in, and the redaction is part of the signed record.
A5
The platform reports what the data indicates. The decision stays with the person who has to defend it.
A6
What was said and what happened. Never what somebody was trying to do.
A7
Every AI-produced line is marked as AI-produced, in the metadata and on the surface a person reads.
A8
Blacklists, discriminatory criteria, routes around human review. The platform declines and says why.
[03 / ENFORCEMENT]
When a rule fires, the refusal is itself a signed event. You can export it, verify it offline, and show a regulator that the gate held.
The full signing story lives on the audit chain page. The platform overview sits on REVEAL · the platform.
[04 / CONSOLE]
Open the console.Same instance our clients run. No login, no sales call.