Enabling Spotter

To enable Spotter search and other AI features for users in your organization, you must first enable AI features from the Admin portal, add Spotter privileges to the specific users or groups, and enable Spotter on your desired Model(s).

Turn on AI features for your instance

To turn on AI features for your instance, follow these steps:

  1. Navigate to the Admin tab. If you have multiple Orgs in your instance, select All Orgs.

  2. Select ThoughtSpot AI under the Application settings section in the left side menu.

    Edit Ask AI section

  3. From the Choose Spotter experience section, you can choose either Spotter or Sage & Ask Sage.

  4. From the Choose Spotter answer experience section, you can choose the answer experience as Spotter agent or Classic experience. You can also set Spotter Agent as on by default, if it has been enabled. For information about the Spotter answer experience, see Spotter answer experience behavior.

    Set Answer Why questions to Enabled to facilitate asking questions about why your data has changed, and to receive natural language summaries of key change drivers.

    Set Enable add to coaching from chat to Enabled to allow users to add reference question and business term interpretations to Spotter without leaving their conversation.

  5. From the Choose where your users can access Spotter section, you can enable Spotter on the home page and left navigation, and on Liveboards.

    If you select Spotter under Choose Spotter experience, but do not enable Spotter on the home page or on Liveboards, users on your instance will not be able to access Spotter.

  6. From the Other AI features section, you can enable the following features:

    AI Sample Questions

    Enable this feature to see data set-specific suggested searches on the Home page. For more information, see AI-suggested searches.

    AI Narratives for SpotIQ

    Also known as AI Highlights, this feature provides you with quick insights on how top metrics changed in your Liveboard. For more information, see AI Highlights.

    AI Highlights on Liveboard

    Enables an AI Highlights button that initiates SpotIQ to produce AI highlights based on the data in the Liveboard.

  7. From the LLM Configuration section, you can choose which provider to use for ThoughtSpot’s AI features. We currently support Azure OpenAI, and Google Gemini (Early Access).

  8. From the Data Models section, you can disable AI search on all data models. To enable AI on a data model, you need to access the individual Model from the Data workspace and turn it on.

  9. From the User privileges section, you can click to be directed to Users and Groups in order to assign privileges.

    Note that users or groups must have Can preview thoughtspot sage permissions to access Spotter.

  10. To revert to the previous settings, click Reset configuration.

Enable Spotter for a user group

To give Spotter privileges to a certain user group, follow these steps:

  1. Navigate to the Admin panel, either by clicking on the Admin button in the top menu or by selecting Admin from the app switcher icon.

  2. If your instance has Orgs enabled, navigate to the Primary org tab in the left menu and select Groups under User management. Otherwise, select Groups under User management.

  3. Select the name of the desired user group to open the Edit group pop-up.

  4. Scroll to the Roles section and select the checkbox next to the can preview thoughtspot sage privilege. Click Review selection and Save.

Enable Spotter for a Liveboard

In order for users to access Spotter on a Liveboard visualization, the underlying data source must be enabled for Spotter. To enable a Model for Spotter, follow these steps:

  1. Navigate to the Data workspace, either by selecting Data in the top menu bar or by selecting Data workspace from the app switcher icon.

  2. Select the Model from the list in the Data workspace.

  3. Click the More icon more menu icon in the upper right corner and select Enable Spotter.

    If Spotter is disabled on a certain Model, users will still see the Spotter icon on visualizations built on any dependent Liveboard. Clicking Spotter will result in an error message prompting users to ask an admin to turn on AI features for the data source.

Choose your Spotter experience

If Spotter Agent has been enabled for your instance, you can choose between Spotter experiences in your user profile. To change your selection, follow these steps:

  1. Navigate to your user profile and scroll to the Experience section.

  2. Click Edit.

  3. For Spotter answer experience, select Spotter agent or Classic experience.

  4. Click Save.

    If you do not find Spotter answer experience in your user profile, contact your ThoughtSpot administrator for assistance.

Spotter answer experience behavior

Refer to the following table to understand the behavior of the Spotter answer experience based on which settings are enabled.

Choose Spotter experience Spotter answer experience Make Spotter agent default experience for all users User experience

Sage & Ask Sage

Spotter Agent

Enabled

We do not recommend this configuration.

Spotter

Classic experience

Disabled

Spotter Agent is disabled on the instance.

Spotter

Spotter Agent

Disabled

Spotter Agent is disabled by default, but you can opt in from your profile page.

Spotter

Spotter Agent

Enabled

All users experience Spotter Agent.

Security

Spotter respects all existing security measures, including row-level security and column-level security.

To access Spotter on a Model, three tiers of permissions are required. Spotter must be enabled for your instance, the Model must have Spotter enabled, and you must have Spotter user privileges.

Users without proper privileges see a pop-up message indicating "Spotter is not enabled for you."

Data shared with LLM

The Answer why questions feature is the only setting that shares data with the underlying LLM.

Enabling this feature allows Spotter to send query-related data, such as aggregated metrics and dimension values, to an LLM to identify the data points and generate explanations. Please ensure this complies with your organization’s policies. The data is not utilized for model training, and we maintain a zero-data retention policy with all our LLM providers.

Was this page helpful?
×