Trust center

Boring is the flex.

Compliance here isn’t a policy PDF — it’s code. Every claim on this page maps to a shipped, deterministic mechanism, and you can test most of them yourself on a live demo call.

The mechanisms

Eight gates, all shipped, all deterministic.

Each card names the mechanism that enforces it. Nothing here relies on an AI “remembering” to behave — where it matters, the rail is code.

Disclosure

The AI says it’s an AI. Every call.

The setter opens every call by disclosing that it’s an AI and that the call is recorded. It isn’t a line in a prompt the model might skip — the greeting is a deterministic gate: the call doesn’t proceed without it.

Deterministic greeting gate, enforced in code
Consent

No consent on file, no cold dial.

Cold outbound calls require express written consent recorded against the contact before the dial is allowed. The gate fails closed: missing proof means the call never happens — it doesn’t mean “probably fine.”

Fail-closed consent gate on every cold dial
Opt-out

One STOP stops everything.

Text STOP — or ALTO — once, and that person is suppressed across every channel we operate: SMS and voice alike. The vocabulary covers English and Spanish, and the match is deterministic string logic, never an AI’s judgment call about intent.

Cross-channel suppression · EN + ES vocabulary · deterministic
Quiet hours

8 a.m. to 9 p.m., the lead’s clock.

Every consumer touch — call, text, or email — is clamped to the federal TCPA 8 a.m.–9 p.m. window in the lead’s local timezone. If we can’t determine the timezone, the touch waits: the clamp fails closed, it never guesses.

Hard clamp to the 8–21 TCPA floor · fail-closed on unknown TZ
Canada

Canadian numbers are never dialed.

CASL, Canada’s anti-spam law, has a stricter consent regime than the U.S. Rather than approximate it, we don’t dial Canadian numbers at all. The gate is a hard block, not a setting.

CASL posture: hard block on Canadian numbers
Quality control

A second AI grades every call.

100% of calls — not a sample — are graded by a second, independent AI for compliance and quality. Some failures are automatic: quoting a price, or claiming to be human, fails the call on the spot and flags it for review.

100% coverage · auto-fail gates: price quotes, human claims
Your data

Your data is yours. Also after you leave.

Clients can export their full data at any time — contacts, calls, outcomes. And your do-not-call list survives churn: a person who opted out stays opted out even after you stop being a client. Compliance isn’t a subscription feature.

Full export any time · DNC list survives churn
Recording

Recording is disclosed up front.

Calls are recorded for the QC loop — that’s how the second AI grades them — and the recording disclosure rides the same deterministic greeting gate as the AI disclosure. Nobody is recorded without hearing it first.

Disclosure rides the same code-enforced greeting gate
What you won’t find here

No invented badges.

We don’t wear badges we haven’t earned. There’s no SOC 2 report or HIPAA attestation to show you today — so you won’t see those logos on this page. And if a vendor tells you they’re “TCPA certified,” walk away: no such certification exists — no regulator issues one. What we have instead is the list above — mechanisms you can test on a live call, starting with the demo on the homepage.

Why “boring”? Because the exciting version of this industry is the one getting sued. A setter that discloses itself every time, refuses unconsented dials, honors one texted STOP everywhere, and gets every call graded isn’t flashy — it’s a precision instrument. Boring is the flex.
See it on yourself — the demo call is the product.