Pricing / operating risk

The cost of oversight is smaller than the cost of an apology.

Flat-rate subscription for the rulebook, scoring, and response layer around your customer-facing AI.

No per-call tax · no inference lock-in

For a live agent

Starter
The essential audit layer for one customer-facing AI workflow.
$1,000/ mo
  • Up to 25k scored turns / mo
  • Core hallucination and PII checks
  • Weekly audit report in PDF + JSON
  • CX-lead email digest
Talk to sales
Most picked

For a growing surface area

Growth
A broader rulebook and real-time escalation for teams scaling AI support.
$2,500/ mo
  • Up to 100k scored turns / mo
  • Custom rules per customer rulebook
  • Real-time Slack, PagerDuty, and email escalation
  • EU AI Act, Colorado, and NYC LL 144 gate sets
  • Quarterly legal-readiness memo
Request an audit slot

For regulated scale

Enterprise
A dedicated operating layer for multiple agents, regions, and rulebooks.
Customscoped to your risk profile
  • 250k+ scored turns / mo
  • Joint rulebook reviews with counsel
  • Dedicated success engineer and office hours
  • Quarterly adversarial test-set refresh
  • Annual bias-audit support

A small sample of the work

What Voxguard catches before it becomes a customer story.

Flagged output

Hallucination · tone drift

Customer-facing response · redacted sample

“I checked your account and can confirm the requested change is already approved. You'll see it reflected shortly.”

[customer reference redacted] · [policy source absent] · [destination redacted]

The answer sounds reassuring, but it asserts an account action without evidence and promises an outcome in the brand's most absolute voice.

Rule

Unsupported account claim + certainty beyond the evidence.

Follow-up action

Response blocked, a safer rewrite queued, and a source-review task opened for the rulebook owner.

Explore the risk library

Every plan starts with a rulebook review and a scoped rollout.

Built for ops · legal · CX