Updated as of Version 26.10.
Ready to upgrade? Contact PeopleFluent Support.

AI Features Set Up

Beyond the AI Assistant, there are several other areas in the application that utilize AI to suggest improvements, generate images and provide translations. These features include:

  • Suggest Course Description Improvements: Enter a course description and request that AI suggest improvements to the description text.
  • Generate a Catalog Image: Request AI to generate a catalog image for a learning module based on the module title and description.
  • Language Bundle Translations: Request AI automatically generate translated fields when language bundles are added to objects that use them. For example, if you add a French language bundle to a learning module, you can request for fields within the module to be automatically translated to French.
  • Language String Translations: Request AI automatically translate individual text labels (strings) for a specific language.

The feature must be enabled and at least one AI model must be configured. You can have multiple AI models and assign target audiences to each model. 

Enable the AI Feature and Configure the Settings

  1. Go to Manage Center > System > General Settings > System Configuration.
  2. In the Select a Category field select System.
  3. Locate the Enable AI field and click the checkbox.
  4. Configure the AI settings:
Field Description
AI Connect Timeout (ms) Specifies the maximum time, in milliseconds, allowed to establish a connection to the AI service before timing out.
AI Read Timeout (ms) Specifies the maximum time, in milliseconds, allowed to wait for a response from the AI service after a connection has been established.
AI Max Retries Sets the maximum number of times the system will automatically attempt to re-send a failed request to the AI service before returning an error.
AI Retry Backoff (ms) Sets the initial delay, in milliseconds, the system waits before making a retry attempt following a failed request to the AI service.
Maximum AI requests per user per minute Sets the maximum number of AI requests an individual user can submit within a rolling one-minute window.
Maximum concurrent AI requests Limits how many AI requests can be processed concurrently across the system.
Maximum AI prompt characters Limits the number of characters allowed in an AI request prompt.
Maximum AI request payload bytes Limits the total size of an AI request payload, including prompt, metadata, artifacts and context.
Maximum AI request artifact count Limits the number of artifacts that can be attached to one AI request.
Maximum AI request artifact bytes Limits the size of each artifact attached to an AI request.
Maximum AI request input tokens Limits the maximum estimated input tokens allowed for each AI request. The estimate includes prompt text, context fields, and INLINE_TEXT artifact content.
Default AI conversation memory policy

Defines the default conversation memory policy to apply when an AI request does not specify one. Available options:

  • None: No conversation history is retained or sent.
  • Ephemeral: Uses short-lived memory scoped to the current request or session.
  • Conversation: Retains memory for the duration of one conversation.
  • Persisted: Stores long-term memory for reuse across sessions.

Role Access Controls

There are two role access control permissions related to the AI features:

  • AI Model Setup allows role users to view and configure AI models.
  • AI Features allows role users to access and use the available AI features.

For information about setting role permissions, please see Create & Manage Roles.

AI Model Setup

Add or Edit a Model

  1. Go to Manage Center > System > General Settings > AI Model Setup.
  2. Select +Create New or open an existing model for editing by clicking its action menu and selecting Edit.
  3. Select an option in the Provider field. We support the following:
  • OpenAI
  • Anthropic
  • Google (Gemini)
  • Google Vertex AI
  • Azure OpenAI
  • Open AI Standard (This refers to the REST API format originally developed for OpenAI’s models, which has become the industry standard for AI inference).
  1. Configure the settings. Each model type has similar setting values, although some are provider specific. In general, you will need the connection settings and API keys. For example, for Azure OpenAI, Azure will provide the base model, the API key and the API endpoint.
  2. Set the target audience. This allows different user groups to utilize different AI models. Options include:
  • User
  • User Group
  • Organization
  • Role
  • User Attributes 1-8
  1. Click Save.

Test the Connection

Ensure that the connection is valid by clicking the Test the Connection button.

Activate a Model

To activate a model for use, open the model for editing and click the Active button.

Set the Model Permissions

You can set permissions for configuring and viewing the configuration on the model level. Click the Permissions button.

  • Read Only gives the selected users view-only access to the module.
  • Unrestricted Access additionally allows the user to edit the configuration.

Delete a Model

To delete a model, locate the appropriate option on the AI Model page. Click the action button and select Delete. Alternatively, open the model and click the Delete button.

Are you a PeopleFluent administrator?
Check out our Help Center or Contact Support.

Was this article helpful?

0 out of 0 found this helpful