Configure sensitivity detection

  • Freigeben Version: Australia
  • Aktualisiert 12. März 2026
  • 1 Minute Lesedauer

    • Configure sensitivity detection in the Now Assist for HR Service Delivery (HRSD) application.

    Vorbereitungen

    Role required:
    • sn_hr_gen_ai.admin

    Warum und wann dieser Vorgang ausgeführt wird

    This capability helps you configure sensitivity detection and filters for the types of sensitive cases such as the harassment complaints, discrimination allegations, workplace violence, safety, employee behavior, and employee personal issues. You can also block Now LLM Service from engaging with these sensitive case types to allow agents to create cases for these sensitive case types.

    Hinweis:

    Configurations are available on the Now Assist Admin Panel settings as part of Now Assist guardian. The sn_hr_gen_ai.admin role can only Edit, Activate and Deactivate skills on Guided setup view only.

    Prozedur

    1. Navigate to the Now Assist Admin Panel and select the Settings tab. Access to Now Assist Admin Panel and Settings tab
    2. Select Now Assist Guardian and select Filters.Filters tab view that lists the existing filters

      Filter configurations include the ability to Edit or Deactivate a filter. Only sn_hr_gen_ai.admin can Edit or Deactivate.Shows the ability to Edit or Deactivate a filter from the list

    3. Select General Details tab to configure the filter details.Shows the ability to add filter type and description details
    4. Select Sample phrases tab to add phrases that apply to filters.Shows the ability to add new and edit existing sample phrases
      Hinweis:
      The ability to edit and update existing filters is available. There is a maximum of 10 phrases that can be generated.
    5. Select Applicability to choose what filters apply to active virtual agents.Shows the ability to add virtual agents to sensitive phrases

    Ergebnisse

    When an employee enters one of the configured sensitive phrases, the virtual agent offers to either redirect the employee to a live agent, create a case, or if the virtual agent incorrectly identifies a sensitive topic, the employee can select to proceed with the interaction.