Skip to content
English
  • There are no suggestions because the search field is empty.

SparkLearn AI Chat for Admins: Empower, Customize, Measure

SparkLearn’s AI Chat gives your learners instant answers — powered entirely by your organization’s own content. As an admin, you’re in control: from setup and customization to measurement and insights. This article covers what you need to know about enabling, managing, and making the most of AI Chat.

How It Works

AI Chat uses the content you’ve already loaded into SparkLearn:

  • Articles
  • PDFs
  • Documents

All of this flows into SparkLearn’s AI engine, creating a single source of truth for your workforce. Learners get reliable, fast answers without digging through multiple systems.

Customization Options

Every organization is unique, and AI Chat reflects that. 

  • Navigate to your System Information page in SparkLearn.
  • Locate AI Chat Settings
  • Set the AI’s identity — choose a name, description and icon (these settings do not have to be changed).
  • Create starter prompts — guide learners on how to begin asking questions. You can choose up to 3 default prompts that will be made available to Learners. 
  • Add a custom placeholder message that will display within the text field for the AI chat. 
  • Add a disclaimer — include safety or compliance messaging if needed.

Measuring Impact

AI Chat isn’t a black box — every interaction is measurable via the AI Chat metrics Dashboard.

Admins can see:

    • Usage volume (overall activity)
    • Feedback (thumbs up/down)
    • The most active users 
    • Questions that returned no response from the AI engine
    • The number of Source documents per response 
    • The most referenced content 
    • Click-throughs to source content
  • Usage over time
  • Topics of conversations
  • Popular usage times
  • Chat error rate

These insights help you:

  • Spot gaps in training materials

  • Monitor effectiveness of your content

  • Understand what your workforce really needs to know

All interactions are saved as xAPI statements, if you have access to Veracity LRS you can dive deeper into the statistics as needed.