ZenotiHelp

Restricted content filtering

ZTZenoti Docs Team Updated April 10, 2026 1 min readGuide

SmartBot monitors conversations for restricted words and automatically protects personally identifiable information (PII) in analytics data, helping administrators maintain conversation safety and data privacy across all channels.

Audience: Administrators monitoring conversation safety and data privacy.

Before you begin

  • Restricted content filtering is active by default for all SmartBot conversations. No configuration is required to enable it.
  • Administrators receive alert notifications when restricted content is detected.

How restricted content filtering works

Restricted word detection

SmartBot scans incoming guest messages for restricted or inappropriate words. When a restricted word is detected, SmartBot:

  1. Sends an alert notification to the administrators configured to receive these alerts. The alert includes:
  • The restricted words that were detected.
  • The message content containing the restricted words.
  • The guest's phone number.
  • The organization and center where the conversation took place.
  1. The conversation continues normally β€” SmartBot does not block the guest or terminate the session automatically.

Administrators can use the alert information to review the conversation in the Inbox and take appropriate action, such as initiating an agent handover to a human agent.

PII redaction in analytics

To protect guest privacy, SmartBot automatically redacts personally identifiable information from conversation data used for analytics. The following categories of information are redacted from analytics records:

  • Names
  • Phone numbers
  • OTP codes
  • Service types
  • Dates
  • Times
  • Prices

PII redaction applies to the analytics data pipeline only. Full conversation content (including unredacted messages) remains available for authorized review in the Inbox.

Was this article helpful?
Your feedback helps us improve our documentation.