Controls
Set autonomy levels, approval rules, and tool access so Bizy knows when to act automatically and when to pause for your review.
How Approvals Work
AI encounters trigger condition
Customer requests something that matches an approval rule.
AI pauses and notifies you
You receive a notification with context and proposed action.
You approve, reject, or modify
One tap to approve, or take over the conversation yourself.
AI continues
After approval, AI completes the action seamlessly.
Common Approval Rules
These are popular rules businesses use to maintain control:
Large bookings
Require approval for bookings over a certain value
Same-day bookings
Review bookings requested for the same day
Cancellations
Get notified when customers want to cancel
Complaints
Be alerted when customers express dissatisfaction
Rule Categories
Booking Rules
- High-value bookings
- Same-day appointments
- New customer first booking
- Specific service types
- Bookings with specific team members
Modification Rules
- Cancellation requests
- Reschedule requests
- Time change requests
- Service change requests
Conversation Rules
- Complaint detection
- Refund requests
- Questions AI can't answer
- Sensitive topics
Custom Rules
- VIP customer interactions
- Off-hours messages
- Specific keywords or phrases
- Channel-specific rules
Autonomy modes
Pick a recommended mode that sets autonomy level and thresholds.
Observer (Level 1)
Read-only. Bizy can suggest actions but approvals are required for any change.
Assistant (Level 2)
Low-risk actions can run automatically. Medium and higher-risk actions need approval.
Trusted (Level 3)
Low and medium risk actions can run automatically. High risk actions still require approval.
Autonomous (Level 4)
Most actions run automatically. Critical actions still require approval.
Advanced override: manually set autonomy levels 1-4 when you need custom boundaries.
Tool Access
Tool access gates what Bizy is allowed to do. Use it to limit high-impact actions while still letting the assistant respond quickly.
- Enable booking creation but require approvals for high-value services.
- Allow CRM updates while pausing refunds or cancellations.
- Disable payment tools until a manager reviews the setup.
Monitor Your AI
Use the AI dashboard to review pending approvals, audit activity, and tune rules as you learn what works best for your business.
Regulated Mode
For businesses in healthcare, finance, or other regulated industries, Regulated Mode enforces stricter evidence and approval requirements.
What changes in Regulated Mode
- Higher confidence thresholds required for auto-approval
- All health/medical questions require evidence citations
- Sensitive topics (pricing, treatments) auto-escalate for review
- AI cannot make claims without verified knowledge sources
Best for: MedSpas, dental offices, financial advisors, and any business where incorrect information could cause harm.
Policy Versioning
Every change to your approval rules is tracked with full audit history. See what changed, when it changed, and who made the change.
Change Log - See all previous versions of each rule
Diff View - Compare what changed between versions
Rollback - Restore previous rule configurations if needed
Export - Download audit history for compliance reporting
Messaging Limits & Guardrails
Built-in protections prevent your AI from over-messaging customers or getting stuck in conversation loops.
Loop Protection
AI detects when a conversation is going in circles and automatically pauses for human intervention instead of continuing unhelpful exchanges.
Rate Limiting
Prevents spam by limiting how frequently AI can message the same customer. Configure per-customer daily/hourly limits for outbound messages.
Allowance Management
Set monthly message allowances per campaign type. Track usage and get alerts when approaching limits.