Skip to content
REVENZA

Safety

Autonomous where permitted. Human-controlled where required.

Permissions are not a disclaimer at the bottom of the page. They are the part of the product that makes the rest usable.

Negotiate within approved limits. Escalate when authority is required.

AI may act autonomously — only where you configure it

  • Answer approved questions
  • Collect approved information
  • Qualify using approved criteria
  • Recommend an approved offer
  • Quote a current approved price
  • Offer an authorized term or discount within a defined ceiling
  • Schedule an approved next step
  • Send an eligible, consented follow-up
  • Transfer to an approved human destination

Human approval required

  • Custom price
  • Discount beyond the autonomous ceiling
  • Unusual payment term
  • Contract or scope change
  • Large, high-risk, sensitive or regulated opportunity
  • Refund, complaint or exception
  • Unavailable or uncertain information
  • Any action the business has reserved for a person

Never permitted

  • Invent a price, product, service, offer, policy, capability or availability
  • Exceed authority
  • Make unauthorized contractual commitments
  • Fabricate scarcity or urgency
  • Make a medical diagnosis, prescribe, evaluate emergencies, or replace licensed professional judgement
  • Bypass consent, suppression, quiet hours or opt-out
  • Promise guaranteed revenue, conversion, savings or customer outcomes
  • Conceal AI use where disclosure is required

Autonomous where permitted. Human-controlled where required.

Close within your rules.

Commercial permissions
  • Approved catalogue18 services · 4 packagesAI may act
  • Trusted price sourcePrice list — currentAI may act
  • Quotation rulesApproved format · 14-day validityAI may act
  • Autonomous discount ceilingUp to 5%AI may act
  • Beyond 5% / custom scopeCommercial lead approvesHuman approval required
  • Refunds and complaintsBranch manager approvesHuman approval required
  • Inventing a price or offerNeverNever permitted
  • Guaranteeing a resultNeverNever permitted
  • Permitted channelsWebsite · Zalo · WhatsApp · VoiceAI may act
  • Consent and quiet hours08:00–20:00 · opt-out honouredAI may act
Live handoff
  1. Authority limit reached
  2. Context package assembled
  3. Named approver notified
  4. Human replies with decision
EscalatedHuman approvalOutcome
The person receives the transcript, the opportunity and the requested term.

What we can prove today

  • Original product demonstrations you can inspect on this site
  • A written permission model with three explicit levels
  • Realistic fictional examples, labelled as simulations
  • A measurable pilot structure with your own baseline
  • Best-fit and poor-fit guidance, including when to say no
  • A sample assessment result before you give us anything

We do not publish customer names, logos or performance figures yet. What you can inspect is the process, the permissions and the limitations.

Practical questions

What happens when a customer wants a person?
It hands over, with the transcript and the opportunity attached. Continuing to negotiate after a handoff request is prohibited.

Build the AI workforce behind your next closed deal.

View standard pricing or start with a clearly scoped Pilot.