Prerequisites
- Status: You must have an active AI Agent created in AI Studio.
⚠️ Important: FAQ Skillset Deprecation Please note that the FAQ skillset will be deprecated soon. We recommend moving your existing FAQ files to the new Knowledge Base to ensure continued functionality.
Why use Skillsets?
Skillsets determine the functional actions your AI Agent can perform. Unlike a standard text-based bot, an agent with skillsets enabled can actively capture leads, look up shipping data, and process sales directly in the chat.Configuration steps
1
Access the Skillset menu
- Navigate to your active AI Agent configuration.
- Scroll down to the Skillset section.
-
Toggle On the skills you wish to enable.
2
Enable Data Collection
Use this skill to gather customer information for leads or retargeting.
1. Select Primary Data
1. Select Primary Data
Select the checkboxes for the specific data you need (e.g., Email, Name).
2. Add Instructions
2. Add Instructions
Type specific instructions in the data field to guide how the AI Agent asks for the info.
3. Extra & Custom Fields
3. Extra & Custom Fields
In addition to primary data, you can enable extra fields or create custom ones (like “Shoe Size”) in the Customers menu.
To add a custom field:
To add a custom field:3
Follow-up & Auto-resolve
Automate conversation lifecycles by nudging unresponsive users and automatically closing inactive tickets.
1. Follow up (Re-engagement)
1. Follow up (Re-engagement)
Goal: Re-engage a customer who has stopped responding.
- Enable: Toggle the switch in the top right.
- Timer: Set the duration (e.g.,
1 Hours) to wait before sending a nudge. - Message: Customize the text (e.g., “Hi, do you have any other questions?”).
2. Auto-resolve (Closure)
2. Auto-resolve (Closure)
Goal: Keep your ticket queue clean by closing abandoned conversations.
- Enable: Toggle the switch in the top right.
- Timer: Set the duration (e.g.,
1 Hours) to wait after the last chat before automatically resolving the ticket. - Closure Message: Optional text sent when the ticket closes (e.g., “What would be your suggestion for this AI Agent?”).