Skip to main content

Reinforcement agent

  • Last updated:

Introduction

The Reinforcement agent turns course completion into an ongoing learning experience by automatically scheduling follow-up practice sessions and delivering in-platform notifications when sessions are ready. The Reinforcement agent is proactive: it runs automatically in the background based on course completion triggers and cannot be initiated manually from the agent menu. Superadmins and Power Users have the ability to manually trigger reinforcement sessions for testing purposes in the course agents configuration menu. To ensure all features of the reinforcement agent work for testing purposes, make sure the Superadmin or Power User you are testing with is actively enrolled in the relevant course.

Activation and configuration

Superadmins and course administrators manage the Reinforcement agent through a combination of platform and course-level settings:

  • Global enablement: Superadmins enable the agent in AgentHub and configure its global branding (icon, color, title, and description). Ensure the “Search and discovery” AI assistant capability is enabled for all users in the Artificial intelligence control panel.
  • Course-level opt-in: Administrators with required permissions explicitly enable reinforcement on a per-course basis within the course settings ("Agents" tab).
  • Cadence configuration: Set the reinforcement schedule and session frequency per course to control how practice moments are spaced out over time.
  • Content eligibility: The agent can only use course content that has a text transcript. Make sure the course training materials are supported content types for text extraction and transcript generation. If a course has no transcript, a banner appears in the Agents tab of the course settings. If you just added new materials, transcript generation can take a few minutes.

Note: The first message the Reinforcement agent sends in a session is always in English, and Superadmins cannot change it at this time. This is expected to be resolved in a future release, when the language of this message will be based on each user's preferred language.

Capabilities

  • Trigger reinforcement automatically on course completion
  • Dynamically adjust difficulty and content per user, based on the interaction history
  • Build progressively across sessions without repeating covered ground
  • Deliver interactive reinforcement activities with contextual feedback

Use cases

  • New-hire onboarding retention: Automatically space out policy, culture, and operational check-ins across a new hire’s first 30, 60, or 90 days to increase long-term recall without manual manager follow-up.
  • Sales and product enablement: Keep product differentiators, positioning, and objection-handling talk tracks top of mind after a major launch or enablement push.
  • Frontline and compliance training: Reinforce critical safety protocols or regulatory standards through short, digestible practice moments delivered directly in the platform.

Best practices

  • Enable selectively: Focus reinforcement on high-impact courses with substantive content where long-term retention directly affects job performance.
  • Align cadence with course depth: Match session frequency to the complexity of the material, avoid long reinforcement sequences for lightweight or simple courses.
  • Verify material readiness: Before enabling the agent for the specific course, check the Agents tab in the course settings for the transcript availability banner, and test the agent in preview. If the course has no transcript, the preview returns an error. Reinforcement sessions never start for a course without a transcript, so learners won't receive practice activities until it is available.
  • Use for practice, not evaluation: Position reinforcement sessions as supportive learning practice rather than a replacement for formal course assessments or certifications.