System Models, tables, and views

Most of the monitoring information in ThoughtSpot’s system Liveboards comes from system Models, tables, and views that administrators can view, but not modify. The underlying tables are protected system tables that you cannot access directly. However, administrators can create new, custom monitoring reports from the Models and views.

List the system Models, tables, and views

To list the system Models, tables, and views:

Go to the Data workspace tab and click Data objects.
Choose All.
Enter System User in the search field.

Description of system Models, tables, and views

Credits Purchased: Contains data on the number of credits for consumption-based pricing your cluster has, and their expiration date. Used for the Credit Usage Liveboard.
Sample Retail - Apparel: Contains sample retail data.
Credit Usage Model: Tracks the credits for consumption-based pricing your cluster has consumed. Used for the Credit Usage Liveboard.
Discover Monitoring Data: This Model contains data on how users are searching for existing Answers and Liveboards. It contains information on what users are searching for, where users are successful and where they are not. This Model contains data from the MetricsMonitoring table, and is used for the Spotter Conversations Liveboard.
Product Usage: Contains usage data for the product. Specifies what existing Models, tables, and views users search on and create objects from, and what those objects are. Specifies what actions users complete in the product. You can also use this Model to view the underlying data sources for any objects, and any object’s dependents.
MetricsMonitoring: This table contains data on how users are searching for existing Answers and Liveboards. It contains information on what users are searching for, where users are successful and where they are not, based on click count and position. This table is a source for the Discover Monitoring Data Model.
TS: Entitlement Tracking: Sources data from Snowflake via a secure view and refreshes every 24 hours to track entitlements for the products purchased by customers. This system Model is automatically provisioned on all ThoughtSpot Cloud instances and shared across all Orgs on the instance by default.
TS: Object Tracking Beta: Contains creation date, creator, last modification date, and last modifier for any supported Liveboard, Model, and Answer. Administrators can use this Model to build custom monitoring Liveboards that show who created or modified objects and when.
TS: Login Activity Tracking: This Model shows how users access the system. It contains username and ID, Org name and ID, login timestamp, whether the login was successful, HTTP Response Code, and login success rate.
TS: Provisioned Users: Contains data related to the current status of the user (active, inactive, or deleted), Org ID and Org name as in ThoughtSpot, user ID, and the number of provisioned users.
TS: AI and BI Stats: Contains query performance metrics for each query made against an external database. It is based on Query Stats, derived from Data Manager. It offers a stable and trusted alternative to other data sources. This data source is already used for billing and has demonstrated reliability. This Model is enabled by default. For more information, see AI and BI Stats and TS: AI and BI Stats.
TS: BI Server: Contains data related to the systems associated with underlying the ThoughtSpot BI server. This includes database latency, browser clients, size of responses, and more.
TS: External Table Current Row Count: Contains data related to the row count for external tables, or tables loaded through Connections.
TS: Rows of un-cached tables for each connection: Contains data related to the sum of rows count for each un-cached table in a connection.

TS: External Table Info: Contains data related to external tables, or tables loaded through Connections.
TS: Daily Row Count External Table: Contains Connections data, stored on a daily basis: the number of rows in your external database, and whether each external table is cached or not.
TS: Query Stats: Contains data related to billable query consumption for customers on the query-based consumption pricing Model. ThoughtSpot distinguishes between regular query consumption and Spotter queries. Used for the Billable Query Stats Liveboard. Spotter queries do not appear in a separate visualization on the Billable Query Stats Liveboard.

How the system Models fit together

ThoughtSpot ships a set of read-only system Models that expose the platform’s own metadata and usage telemetry inside ThoughtSpot itself. You search them, build Liveboards on them, and pull from them through the same Data API you use for any other Model. Administrators of the Primary Org see data across all Orgs; an Org administrator sees only their own Org. Nothing needs to be installed: the Models are provisioned automatically on ThoughtSpot Cloud.

To let a non-admin analyst use a system Model, share both the Model and any Liveboard built on it.

Which Model answers which question

You want to know… Use this Why

You want to know…	Use this	Why
How many users am I paying for, in total and per Org?	TS: Provisioned Users	Purpose-built license count. Org-aware. Excludes deleted users through the Is Active flag.
When did a user last log in, and to which Org?	TS: Login Activity Tracking	Records every login attempt with Org, timestamp, and HTTP response. The only Model that captures logins rather than queries.
Who is inactive in a given Org for 30, 60, or 90 days?	TS: Login Activity Tracking (primary) + TS: AI and BI Stats (secondary)	Login recency answers "has not signed in". Query recency (AI and BI Stats) answers "has not used content". Use login for deprovisioning; use query for adoption.
Who created or last changed this Liveboard, and is anyone still using it?	TS: Object Tracking	Joins object metadata (author, created, modified) to the query telemetry behind AI and BI Stats.
Which content can be archived?	TS: Object Tracking	Filter on Object Modified Time and last User Action Time to find stale, unused objects per Org.
What did we buy, and how much is consumed?	TS: Entitlement Tracking	Sourced from ThoughtSpot’s commercial system every 24 hours.
Adoption, query performance, credits, Spotter usage	TS: AI and BI Stats	See TS: AI and BI Stats and the AI and BI Stats Liveboard.
The definitive list of users and Org membership right now	REST API `POST /api/rest/2.0/users/search`	Live metadata, never lagged. Reconcile Model output against this before taking action on an account.

How many users am I paying for, in total and per Org?

TS: Provisioned Users

Purpose-built license count. Org-aware. Excludes deleted users through the Is Active flag.

When did a user last log in, and to which Org?

TS: Login Activity Tracking

Records every login attempt with Org, timestamp, and HTTP response. The only Model that captures logins rather than queries.

Who is inactive in a given Org for 30, 60, or 90 days?

TS: Login Activity Tracking (primary) + TS: AI and BI Stats (secondary)

Login recency answers "has not signed in". Query recency (AI and BI Stats) answers "has not used content". Use login for deprovisioning; use query for adoption.

Who created or last changed this Liveboard, and is anyone still using it?

TS: Object Tracking

Joins object metadata (author, created, modified) to the query telemetry behind AI and BI Stats.

Which content can be archived?

TS: Object Tracking

Filter on Object Modified Time and last User Action Time to find stale, unused objects per Org.

What did we buy, and how much is consumed?

TS: Entitlement Tracking

Sourced from ThoughtSpot’s commercial system every 24 hours.

Adoption, query performance, credits, Spotter usage

TS: AI and BI Stats

See TS: AI and BI Stats and the AI and BI Stats Liveboard.

The definitive list of users and Org membership right now

REST API POST /api/rest/2.0/users/search

Live metadata, never lagged. Reconcile Model output against this before taking action on an account.

Refresh cadence

The system Models are populated by a telemetry pipeline, not read live from the metadata store. Plan around these cadences:

Model	Refresh	Practical implication
TS: Provisioned Users	Every 2–3 days	A user created or deleted today may not appear for up to three days. Do not use it for same-day license reconciliation; use it for trend and per-Org reporting.
TS: Login Activity Tracking	Daily (previous day, D-1)	Today’s logins appear tomorrow. Set inactivity thresholds at least a day longer than your policy to allow for this.
TS: Object Tracking	Every 2–3 days	Same lag as Provisioned Users for the author and modified metadata; the usage columns follow the daily query-stats refresh.
TS: Entitlement Tracking	Every 24 hours	Consumption to date is as of the previous day.
TS: AI and BI Stats	D-1 to D-2	See TS: AI and BI Stats.

Model

Refresh

Practical implication

TS: Provisioned Users

Every 2–3 days

A user created or deleted today may not appear for up to three days. Do not use it for same-day license reconciliation; use it for trend and per-Org reporting.

TS: Login Activity Tracking

Daily (previous day, D-1)

Today’s logins appear tomorrow. Set inactivity thresholds at least a day longer than your policy to allow for this.

TS: Object Tracking

Every 2–3 days

Same lag as Provisioned Users for the author and modified metadata; the usage columns follow the daily query-stats refresh.

TS: Entitlement Tracking

Every 24 hours

Consumption to date is as of the previous day.

TS: AI and BI Stats

D-1 to D-2

See TS: AI and BI Stats.

Source of truth

Recommended rule for governance automation: Decide with the Model, act with the API.

Use TS: Login Activity Tracking and TS: Provisioned Users to identify candidate users, then confirm each one against POST /api/rest/2.0/users/search before removing them from an Org. The API reflects the current state; the Models reflect the state as of the last refresh.

Treat differences between the two as expected lag, not as data quality failures, unless the gap exceeds the refresh window.

Filtering rules that explain "missing" users

The pre-built Inactive Users and Active Users visualizations exclude tsadmin, system, and internal ThoughtSpot-domain users by design. Build your own Answer if you need those accounts included.
The Provisioned Users KPI on the AI and BI Stats Liveboard does not respond to the Org filter. For per-Org counts, search TS: Provisioned Users directly.
Users who have never run a query do not appear in TS: AI and BI Stats or TS: Object Tracking. They do appear in TS: Provisioned Users and, once they log in, in TS: Login Activity Tracking.
A user who belongs to several Orgs appears once per Org in TS: Provisioned Users and TS: Login Activity Tracking. Always group by User ID and Org.

Build a governance reporting pack

The user provisioning, access, and content governance Models are designed to be combined. A practical monthly governance Liveboard for a platform team could contain the following visualizations:

Visualization Model Search or definition

Visualization	Model	Search or definition
License consumption vs entitlement	TS: Provisioned Users + TS: Entitlement Tracking	`provisioned users` in one visualization, alongside `quantity purchased quantity consumed product offering` in another
Provisioned users per Org	TS: Provisioned Users	`provisioned users org`, sorted descending
Monthly logged-in users per Org	TS: Login Activity Tracking	`unique count user id org timestamp monthly http response code = 200`
Inactive users (candidates) per Org	TS: Login Activity Tracking	`user id user max timestamp org`, filtered in your application at 30, 60, or 90 days
Login success rate trend	TS: Login Activity Tracking	`login success rate org timestamp weekly`
Stale content per Org	TS: Object Tracking	`object name object type object author name object modified time org object modified time < 180 days ago`
Top used Liveboards per Org	TS: Object Tracking	`object name org impressions active users last 30 days object type = 'LIVEBOARD'`
Adoption and performance	TS: AI and BI Stats	See TS: AI and BI Stats

License consumption vs entitlement

TS: Provisioned Users + TS: Entitlement Tracking

provisioned users in one visualization, alongside quantity purchased quantity consumed product offering in another

Provisioned users per Org

TS: Provisioned Users

provisioned users org, sorted descending

Monthly logged-in users per Org

TS: Login Activity Tracking

unique count user id org timestamp monthly http response code = 200

Inactive users (candidates) per Org

TS: Login Activity Tracking

user id user max timestamp org, filtered in your application at 30, 60, or 90 days

TS: Login Activity Tracking

login success rate org timestamp weekly

Stale content per Org

TS: Object Tracking

object name object type object author name object modified time org object modified time < 180 days ago

Top used Liveboards per Org

TS: Object Tracking

object name org impressions active users last 30 days object type = 'LIVEBOARD'

Adoption and performance

TS: AI and BI Stats

See TS: AI and BI Stats

Data API access

You can run every search above programmatically with the Data API (POST /api/rest/2.0/searchdata), passing the Model’s GUID as logical_table_identifier and the search string as query_string. This is the recommended way to feed a deprovisioning or archiving workflow: it is independent of Liveboard layout and returns structured JSON.

Use a service account with administrator rights on the Orgs you are reporting on, and remember that results reflect the last refresh of each Model.

Get help

If a system Model appears empty, returns data older than the refresh windows, or a column described in this documentation is not visible on your cluster, contact ThoughtSpot Support and include:

The Model name and GUID
An example search that demonstrates the issue

Column-level descriptions are also available inside ThoughtSpot by hovering over a column in the Model.

We highly recommend that you register for the free Monitoring Performance course in ThoughtSpot University that covers the information in this article, and watch the System Models lesson.

See other training resources at ThoughtSpot University.