It's possible your conversational AI application will handle personally identifiable information (PII) whilst interacting with users or customers.

UI Builder ships with AI-powered capabilities that can automatically identify and redact sensitive information in utterances. Some example of the types of PII the AI capabilities will redact include, but are not limited to; names of people, job types held by people, phone numbers, email or postal addresses and more.

<aside> 💡 Find more examples of the types of PII your conversational AI application can redact here.

</aside>

Three levels of PII redaction are supported:

1. Allow All PII recording - no filtering. All user input is logged in backend analytics.
2. Filter PII recording - PII data in utterances is redacted in backend analytics.
3. Block All PII recording - no utterances are logged in backend analytics.

You can use this capability to ensure your conversational AI application adheres to compliance or legislative processes. All 3 levels are added using the designer canvas in UI Builder like any other action.

The following links will take you to guides that show you how to configure and verify that PII redaction has been activated in you conversational AI application.

Allow All PII Recording

Filter PII Recording

Custom Filters for PII Recording

Block All PII Recording