AI Settings

The AI settings page allows admins to configure system wide AI / machine learning options.

Enable Generative AI and Natural Language Operations

  • Select the Enable Generative AI and Natural Language Options checkbox to enable the Chatbot, Generative AI, and the Chatbot Speech to Text features. This top-level checkbox must be selected to enable any of the AI features in Discover, Present, Publish, and so on.

Enable Chatbot Features

  • Default Chatbot Engine: Select the type of LLM engine to use for Pyramid's Chatbot. This can be Pyramid Internal or any of the LLMs that have been created in LLM Manager.

Note: By default, this engine will be used by the Chatbot in Discover, Present, and Publish. You can override this default selection on a per model basis in the NLQ panel for that model (Data Model Management).

Enable Generative AI Features

The Generative AI Integration allows users to generate dashboards, queries, images, scripts, excel formulas, translations, choose colors, and more using AI:

  • Select the Enable Generative AI Features checkbox to enable Generative AI Features.
  • Default LLM Engine: Select the default LLM engine to use for Generative AI. The selected LLM engine is used to generate images (in Present and Publish), scripts, excel formulas, translations, choose colors, and to power the GenAI nodes in a Data Flow (used to translate, analyze sentiment, and run general prompts). You can select any of the LLMs that have been created in LLM Manager.

Note: Unlike the Chatbot engine, you cannot override the engine used to power Generative AI Features on a per-model basis, nor are you offered a Pyramid Internal option.

Enable Chatbot Speech to Text

  • Select the EnableChatbot Speech to Text checkbox to enable speech to text when using the Chatbot.
  • Provider: Select the provider of the engine to be used to convert speech to text. This can be one of Google or OpenAI-Whisper.

Google Options

  • Upload Service Account JSON file: Click on the button to upload the required JSON file.
  • Recognizer Name: Select the required name for the Google service from the drop down options that will be populated after selecting a valid JSON file.
  • Advanced Settings
    • Voice Gender: Select the gender of the voice to be used
    • Voice Selection: Select the voice to be used
    • Voice Speed: Adjust the voice speed
    • Voice Pitch: Adjust the voice pitch
    • Voice Volume: Adjust the voice volume
  • Click Test to ensure you are satisfied with your selected gender, speed, pitch, and volume.

OpenAI-Whisper Options

  • OpenAI API Key: Enter the required API key
  • Advanced Settings
    • Voice: Select the desired voice type to be used
    • Voice Speed: Adjust the voice speed
  • Click Test to ensure you are satisfied with your selected voice and speed.

Note: The Chatbot's speech to text feature is only available with Enterprise licensing.