Custom PII Recognizers

As an administrator, you can create and manage organization-specific PII recognizers to enhance data protection. These custom recognizers allow you to detect and mask sensitive information beyond the default set provided out-of-the-box (OOTB) in Luma VA.

Creating a Custom PII Recognizer

To create a new custom PII recognizer, follow these steps:

1. Navigate to Tenant → Settings → PII Recognizer.

2. Click on Create PII Recognizer.

   Open

   

3. Configure the following attributes:

   • Name: A unique name to identify the custom PII recognizer.

   • Status: Defines whether the recognizer is active or inactive. Only active recognizers will detect and mask PII.

   • Entity Name: A system-assigned unique identifier used for internal reference.

   • Mask in Bot Responses: A checkbox option to determine whether detected PII should be masked in bot responses during User and Virtual agent conversations.

   • Description: A brief explanation of the recognizer’s purpose, including details on the type of PII it identifies.

   • Regular Expression: A predefined pattern used to detect specific PII formats, such as email addresses, SSNs, or credit card numbers.

   • Context Words: Additional keywords that help improve detection accuracy by boosting confidence when identifying PII entities. Add word and press enter to add the word.

4. Click on 'Create' to create the recognizer.

Open

Example Custom Recognizer Configuration

Here are a few examples of custom recognizers that you can create:

Editing a Custom PII Recognizer

Once a Custom PII recognizer is created, administrators can modify its settings as required:

1. Navigate to Tenant → Settings → PII Recognizer.

2. Select the recognizer from the list and click Edit.

3. The following fields can be updated:

   • Status: Enable or disable the recognizer.

   • Mask in Bot Responses: Define if detected PII should be masked in bot responses and conversation history.

   • Description: Update the description to provide more details about the recognizer.

   • Regular Expression: Modify the regex pattern for detection.

   • Context Words: Add or remove context words to refine detection accuracy.

4. Click Save to apply the changes.