Embed AI Search and Analytics

Table of Contents

SDK libraries for embedding Spotter
Feature status and availability in embed mode
Related information

Spotter provides an interactive AI-powered search and conversational analytics experience for ThoughtSpot application users. It empowers users to ask questions in natural language and analyzes data to generate business-ready insights instantly.

ThoughtSpot supports three experience modes: Spotter Classic, Spotter Agent (Spotter 2), and Spotter 3. Spotter 3 supports several new features, agentic analytics, and provides an enhanced user experience.

Version	Supported features
Spotter 3 Early Access	Spotter 3 is the latest and most advanced version of Spotter. Key features and user experience: In addition to the Spotter Classic and Spotter 2 capabilities, Spotter 3 supports the following features and enhancements to improve the AI search and analytics experience: Is fully data-aware and capable of understanding data models and schemas Supports advanced analysis Supports automatic discovery and selection of data models based on user prompts. Supports quick search and deep research modes Supports saving chats and accessing chat history Supports third-party MCP connectors Available from 26.2.0.cl and later. To enable Spotter 3 on your instance, contact your ThoughtSpot administrator. When to use: Use this version if you want to include advanced AI search, research, and analytics capabilities in your ThoughtSpot embedded app.
Spotter 2	Also known as Spotter Agent, it’s the second version of Spotter with the following capabilities: Key features and user experience: Agentic AI analytics experience Context-aware analysis Why questions for in-depth analysis Coaching in conversations Data model instructions and guidance Available when the Spotter 2 experience is enabled on the instance. When to use: Use this version if you want to include agentic analytics and richer insights in your ThoughtSpot embedded app.
Spotter Classic	An early version of Spotter. Key features and user experience: Quick search with queries in natural language format Visualization generation Available if the Spotter Classic version is enabled on the instance. When to use: You can use this version if your ThoughtSpot embedded app needs only metadata-level insights and does not require or support agentic workflows.

Version

Supported features

Spotter 3 Early Access

Spotter 3 is the latest and most advanced version of Spotter.

Key features and user experience:
In addition to the Spotter Classic and Spotter 2 capabilities, Spotter 3 supports the following features and enhancements to improve the AI search and analytics experience:

Is fully data-aware and capable of understanding data models and schemas
Supports advanced analysis
Supports automatic discovery and selection of data models based on user prompts.
Supports quick search and deep research modes
Supports saving chats and accessing chat history
Supports third-party MCP connectors

Available from 26.2.0.cl and later. To enable Spotter 3 on your instance, contact your ThoughtSpot administrator.

When to use:
Use this version if you want to include advanced AI search, research, and analytics capabilities in your ThoughtSpot embedded app.

Spotter 2

Also known as Spotter Agent, it’s the second version of Spotter with the following capabilities:

Key features and user experience:

Agentic AI analytics experience
Context-aware analysis
Why questions for in-depth analysis
Coaching in conversations
Data model instructions and guidance

Available when the Spotter 2 experience is enabled on the instance.

When to use:
Use this version if you want to include agentic analytics and richer insights in your ThoughtSpot embedded app.

Spotter Classic

An early version of Spotter.

Key features and user experience:

Quick search with queries in natural language format
Visualization generation

Available if the Spotter Classic version is enabled on the instance.

When to use:
You can use this version if your ThoughtSpot embedded app needs only metadata-level insights and does not require or support agentic workflows.

SDK libraries for embedding Spotter🔗

ThoughtSpot provides the following Visual Embed SDK libraries for embedding Spotter capabilities in your app:

SpotterEmbed
Use this SDK component to embed the full Spotter experience in your app. The features in a Spotter embed vary based on the Spotter version.
SpotterAgentEmbed
Also known as bodyless embedding, SpotterAgentEmbed allows you to integrate Spotter capabilities into your own applications or chatbot, route user questions to ThoughtSpot, and receive structured answers and visualizations. Unlike the full Spotter experience, the "bodyless" embed includes only a prompt panel with no additional buttons or selectors.

Feature status and availability in embed mode🔗

To enable Spotter experience and AI features on your instance, contact your ThoughtSpot administrator.

Feature Supported version, feature availability, and SDK controls

Feature	Supported version, feature availability, and SDK controls
Spotter search and prompt panel for conversational analytics For quick searches with queries in natural language and for generating visualizations.	Supported Spotter version: ✓ Spotter 3 ✓ Spotter 2 ✓ Spotter Classic Feature status: Generally Available (GA) on all ThoughtSpot Cloud instances. Required settings: Spotter feature must be enabled at the instance level. Embed SDK component: If you want to embed the full Spotter experience, use `SpotterEmbed`. To embed only the Spotter search bar with no additional features or standard Spotter user experience, use `SpotterAgentEmbed`. Default state in embed: Enabled by default if the ThoughtSpot instance has Spotter experience enabled.
Localization support Locale settings and language support for Spotter queries and responses.	Supported Spotter version: ✓ Spotter 3 ✓ Spotter 2 ✓ Spotter Classic Feature status: GA. Available on all ThoughtSpot Cloud instances with Spotter feature enabled. Embed SDK component: Use `SpotterEmbed`. The SDK also supports configuring locale settings for `SpotterAgentEmbed`. Required settings: Locale settings can be configured at the user level in the UI or via REST API. Default state in embed: By default, the locale settings configured at the instance level will be used. To override the locale settings in your embed, use the locale property in the SDK.
New chat interface The updated chat interface that provides a more intuitive and interactive analytics experience.	Supported Spotter version: ✓ Spotter 3 x Spotter 2 x Spotter Classic Feature status: Early Access. Available by default on instances that have Spotter 3 enabled. Required settings: Spotter 3 must be enabled at the instance level. Embed SDK component: Use `SpotterEmbed`. Default state in embed: Disabled by default. To enable this feature, set `updatedSpotterChatPrompt` to `true` in the SDK.
Thinking and reasoning UX The thinking and reasoning UX in Spotter conversations shows step-by-step reasoning, planning, and analysis process as Spotter works through a complex question.	Supported Spotter version: ✓ Spotter 3 ✓ Spotter 2 x Spotter Classic Feature status: Early Access. Enabled by default on instances with Spotter 2 or Spotter 3 enabled. Required settings: Spotter 2 or Spotter 3 experience must be enabled at the instance level. Embed SDK component: Use `SpotterEmbed`. Default state in embed: Enabled by default if the feature is enabled at the instance level.
Spotter coaching within conversation sessions The coaching workflow in Spotter conversations.	Supported Spotter version: ✓ Spotter 3 ✓ Spotter 2 x Spotter Classic Feature status: GA in 26.2.0.cl and disabled by default. Embed SDK component: Use `SpotterEmbed`. Required settings: Requires Spotter 2 or Spotter 3 experience to be enabled at the instance level. Default state in embed: Enabled by default in 26.2.0.cl and later versions. Developers can use the `Action.InConversationTraining` action ID to control the visibility of this feature in their embeds.
Data model instructions Global rules that provide data model guidance to the Spotter system and direct how Spotter should interpret queries and construct responses for user prompts.	Supported Spotter version: ✓ Spotter 3 ✓ Spotter 2 x Spotter Classic Feature status: GA. Enabled by default on ThoughtSpot instances with 26.2.0.cl release version or later. Required settings: Spotter 2 or Spotter 3 experience must be enabled at the instance level. Embed SDK component: Use `SpotterEmbed` Default state in embed: Enabled by default.
Deep Research and analysis mode Ability to autonomously perform an in-depth and multi-step analysis on a user’s question.	Supported Spotter version: ✓ Spotter 3 x Spotter 2 x Spotter Classic Feature status: GA. Enabled by default on instances with Spotter 3 enabled. Required settings: Spotter 3 experience must be enabled at the instance level. Embed SDK component: Use `SpotterEmbed` Default state in embed: Available by default if Spotter 3 is enabled on the instance. ThoughtSpot embedded app users can select the deep research mode in the Spotter interface.
Advanced analysis Ability to function as a data-aware analyst, derive context data model, and perform complex research in Research mode for accurate analysis and deeper insights. Supports forecasting, correlation, and query-on-query analysis.	Supported Spotter version: ✓ Spotter 3 x Spotter 2 x Spotter Classic Feature status: Early Access. Enabled by default if Spotter 3 is enabled on the instance. Required settings: Spotter 3 experience must be enabled on the instance. Embed SDK component: Use `SpotterEmbed`. Default state in embed: Available by default if Spotter 3 experience is enabled on the instance.
Why questions Ability to ask questions to automatically generate explanations and identify key drivers behind changes in a metric using change analysis and AI-generated summaries.	Supported Spotter version: ✓ Spotter 3 ✓ Spotter 2 x Spotter Classic Feature status: Early Access. Disabled by default. Required settings: Spotter 2 or Spotter 3 experience must be enabled on the instance. Embed SDK component: Use `SpotterEmbed` Default state in embed mode: Disabled by default. If the feature is enabled on the instance, it will be available in the embed mode.
Auto mode Automatic data source discovery and model selection by Spotter.	Supported Spotter version: ✓ Spotter 3 x Spotter 2 x Spotter Classic Feature status: Early Access. Disabled by default. Required settings: Spotter 3 experience and the Auto-mode feature must be enabled at the instance level. Embed SDK component: Use `SpotterEmbed`. Default state in embed: Disabled by default. To enable this feature, set the `worksheetId` attribute to `_spotter_data_source_discovery_id` in the SDK. When the Auto mode is enabled, the Preview data option in the conversation panel is hidden.
Chat history Saves Spotter conversations and allows reloading chat history with past conversations.	Supported Spotter version: ✓ Spotter 3 x Spotter 2 x Spotter Classic Feature status: Early Access. Enabled by default if Spotter 3 is enabled on the instance. Required settings: Spotter 3 experience and the Chat history feature must be enabled at the instance level. Embed SDK component: Use `SpotterEmbed`. Default state in embed: Disabled by default. To enable this feature in the embed mode, set `enablePastConversationsSidebar` to `true` in the SDK.
Third-party connectors and MCP tools Third-party tools and services that Spotter can connect to and interact with via Model Context Protocol (MCP).	Supported Spotter version: ✓ Spotter 3 x Spotter 2 x Spotter Classic Feature status: Early Access. Enabled by default if Spotter 3 is enabled on the instance. Required settings: Spotter 3 experience and Spotter connectors feature must be enabled on the instance. For connectors to be available in the embed mode, ThoughtSpot administrators must configure Spotter connectors. Embed SDK component: Use `SpotterEmbed`. Default state in embed: Visible by default if the `updatedSpotterChatPrompt` attribute is set to `true`. We do not recommend using this feature in your production environment. You can hide this option using the CSS selectors. For more information, see MCP connectors in Spotter embedding.

Spotter search and prompt panel for conversational analytics

For quick searches with queries in natural language and for generating visualizations.

Supported Spotter version:
✓ Spotter 3
✓ Spotter 2
✓ Spotter Classic

Feature status: Generally Available (GA) on all ThoughtSpot Cloud instances.

Required settings: Spotter feature must be enabled at the instance level.

Embed SDK component: If you want to embed the full Spotter experience, use SpotterEmbed.
To embed only the Spotter search bar with no additional features or standard Spotter user experience, use SpotterAgentEmbed.

Default state in embed: Enabled by default if the ThoughtSpot instance has Spotter experience enabled.

Localization support

Locale settings and language support for Spotter queries and responses.

Supported Spotter version:
✓ Spotter 3
✓ Spotter 2
✓ Spotter Classic

Feature status: GA. Available on all ThoughtSpot Cloud instances with Spotter feature enabled.

Embed SDK component: Use SpotterEmbed. The SDK also supports configuring locale settings for SpotterAgentEmbed.

Required settings: Locale settings can be configured at the user level in the UI or via REST API.

Default state in embed: By default, the locale settings configured at the instance level will be used. To override the locale settings in your embed, use the locale property in the SDK.

New chat interface

The updated chat interface that provides a more intuitive and interactive analytics experience.

Supported Spotter version:
✓ Spotter 3
x Spotter 2
x Spotter Classic

Feature status: Early Access. Available by default on instances that have Spotter 3 enabled.

Required settings: Spotter 3 must be enabled at the instance level.

Embed SDK component: Use SpotterEmbed.

Default state in embed: Disabled by default. To enable this feature, set updatedSpotterChatPrompt to true in the SDK.

Thinking and reasoning UX

The thinking and reasoning UX in Spotter conversations shows step-by-step reasoning, planning, and analysis process as Spotter works through a complex question.

Supported Spotter version:
✓ Spotter 3
✓ Spotter 2
x Spotter Classic

Feature status: Early Access. Enabled by default on instances with Spotter 2 or Spotter 3 enabled. Required settings: Spotter 2 or Spotter 3 experience must be enabled at the instance level.

Embed SDK component: Use SpotterEmbed.

Default state in embed: Enabled by default if the feature is enabled at the instance level.

Spotter coaching within conversation sessions
The coaching workflow in Spotter conversations.

Supported Spotter version:
✓ Spotter 3
✓ Spotter 2
x Spotter Classic

Feature status: GA in 26.2.0.cl and disabled by default.

Embed SDK component: Use SpotterEmbed.

Required settings: Requires Spotter 2 or Spotter 3 experience to be enabled at the instance level. Default state in embed: Enabled by default in 26.2.0.cl and later versions. Developers can use the Action.InConversationTraining action ID to control the visibility of this feature in their embeds.

Data model instructions

Global rules that provide data model guidance to the Spotter system and direct how Spotter should interpret queries and construct responses for user prompts.

Supported Spotter version:
✓ Spotter 3
✓ Spotter 2
x Spotter Classic

Feature status: GA. Enabled by default on ThoughtSpot instances with 26.2.0.cl release version or later.

Required settings: Spotter 2 or Spotter 3 experience must be enabled at the instance level.

Embed SDK component: Use SpotterEmbed

Default state in embed: Enabled by default.

Deep Research and analysis mode

Ability to autonomously perform an in-depth and multi-step analysis on a user’s question.

Supported Spotter version:
✓ Spotter 3
x Spotter 2
x Spotter Classic

Feature status: GA. Enabled by default on instances with Spotter 3 enabled.

Required settings: Spotter 3 experience must be enabled at the instance level.

Embed SDK component: Use SpotterEmbed

Default state in embed: Available by default if Spotter 3 is enabled on the instance. ThoughtSpot embedded app users can select the deep research mode in the Spotter interface.

Advanced analysis

Ability to function as a data-aware analyst, derive context data model, and perform complex research in Research mode for accurate analysis and deeper insights. Supports forecasting, correlation, and query-on-query analysis.

Supported Spotter version:
✓ Spotter 3
x Spotter 2
x Spotter Classic

Feature status: Early Access. Enabled by default if Spotter 3 is enabled on the instance.

Required settings: Spotter 3 experience must be enabled on the instance.

Embed SDK component: Use SpotterEmbed.

Default state in embed: Available by default if Spotter 3 experience is enabled on the instance.

Why questions

Ability to ask questions to automatically generate explanations and identify key drivers behind changes in a metric using change analysis and AI-generated summaries.

Supported Spotter version:
✓ Spotter 3
✓ Spotter 2
x Spotter Classic

Feature status: Early Access. Disabled by default.

Required settings: Spotter 2 or Spotter 3 experience must be enabled on the instance.

Embed SDK component: Use SpotterEmbed

Default state in embed mode: Disabled by default. If the feature is enabled on the instance, it will be available in the embed mode.

Auto mode
Automatic data source discovery and model selection by Spotter.

Supported Spotter version:

✓ Spotter 3
x Spotter 2
x Spotter Classic

Feature status: Early Access. Disabled by default.

Required settings: Spotter 3 experience and the Auto-mode feature must be enabled at the instance level.

Embed SDK component: Use SpotterEmbed.

Default state in embed: Disabled by default. To enable this feature, set the worksheetId attribute to _spotter_data_source_discovery_id in the SDK. When the Auto mode is enabled, the Preview data option in the conversation panel is hidden.

Chat history

Saves Spotter conversations and allows reloading chat history with past conversations.

Supported Spotter version:
✓ Spotter 3
x Spotter 2
x Spotter Classic

Feature status: Early Access. Enabled by default if Spotter 3 is enabled on the instance.

Required settings: Spotter 3 experience and the Chat history feature must be enabled at the instance level.

Embed SDK component: Use SpotterEmbed.

Default state in embed: Disabled by default. To enable this feature in the embed mode, set enablePastConversationsSidebar to true in the SDK.

Third-party connectors and MCP tools
Third-party tools and services that Spotter can connect to and interact with via Model Context Protocol (MCP).

Supported Spotter version:

✓ Spotter 3
x Spotter 2
x Spotter Classic

Feature status: Early Access. Enabled by default if Spotter 3 is enabled on the instance.

Required settings: Spotter 3 experience and Spotter connectors feature must be enabled on the instance. For connectors to be available in the embed mode, ThoughtSpot administrators must configure Spotter connectors.

Embed SDK component: Use SpotterEmbed.

Default state in embed: Visible by default if the updatedSpotterChatPrompt attribute is set to true. We do not recommend using this feature in your production environment. You can hide this option using the CSS selectors. For more information, see MCP connectors in Spotter embedding.

Related information🔗

For more information about Spotter versions and capabilities, see ThoughtSpot Product Documentation.
For information about how to embed Spotter and customize the Spotter interface, see Embed Spotter and SpotterEmbedViewConfig.
For information about bodyless embedding, see Embed Spotter Agent and SpotterAgentEmbedConfig.
For information about the action customization framework to show or hide actions in the embedded view, see Action and Action IDs in the SDK.
For information about event hooks, see Events and app interactions, Embed events, and Host events.
For information about customization style and CSS variables, see style customization framework.