ThoughtSpot Cloud Documentation

May 2023 9.2.0.cl

In this release we introduce our new AI-powered search experience called ThoughtSpot Sage, which makes it easier for you to get answers and insights from your complex data. ThoughtSpot Sage uses a combination of GPT-3, GPT-3.5T and GPT-4, and the large language models (LLM) of the Microsoft Azure OpenAI Service.

Sage features include:

Sage features are disabled by default and available for select customers to try before they reach GA, when they are enabled by default. To enable ThoughtSpot Sage features, you must contact ThoughtSpot Support.

Once enabled, your ThoughtSpot administrator can grant access to the Natural language search and AI-suggested searches features for users in a group by assigning the Can preview ThoughtSpot Sage user privilege to the group. For more information about ThoughtSpot Sage, see ThoughtSpot Sage.

You can now search using natural language to access AI-generated answers to your business questions. To try it out, navigate to the Home page, select the search bar labeled Ask any business question you have and type your question. When you enter your question, you will see the AI-generated answer appear, along with a list of related Answers created by users in your team. For more information, see Natural language search.

You can select the Worksheet to search on by clicking the Worksheet title in the left side of the search bar and selecting a new Worksheet from the dropdown. Worksheets with AI answers enabled display a triangular icon to the left of the Worksheet name: . See Enable or disable AI-generated Answers. This feature must be enabled by ThoughtSpot Support.

The AI-suggested searches feature helps you think of relevant questions to ask of your own data source, by providing examples of searches. This feature is designed to help you get started if you are a new user of Search Data, or if you are working with a new Worksheet. For more information, see AI-suggested searches. This feature must be enabled by ThoughtSpot Support.

Right-click any data point in your Liveboard and easily filter the entire Liveboard by that value. For example, if you right-click California in a geo chart on your Liveboard, and then select Filter, all the other visualizations only show data from California. A visualization for Total sales by city, for example, would only show sales for cities in California. This feature is also called brushing and linking. To try it out, right-click any data point in a Liveboard visualization, and select Filter from the menu that appears. This feature must be enabled by your administrator. For more information about cross filters, see Liveboard cross filters.

Add notes to your Liveboards, so that viewers of your Liveboards can quickly understand how to use the data and visualizations you gathered. Note tiles can provide an overview of a Liveboard or visualization, company branding through images, definitions of terms used in the Liveboard, and so on. To try it out, edit a Liveboard, and select Add note. The Liveboard note tiles feature must be enabled by your administrator. For more information about note tiles, see Liveboard note tiles.

Mandatory Liveboard filters are now GA. Mark certain filters in your Liveboard as mandatory. If you add mandatory filters to a Liveboard, the visualizations will only load after Liveboard viewers select values for the mandatory filters. Use mandatory filters for Liveboards that contain data that may not be relevant for everyone who must view and consume the Liveboard. Mandatory filters also lower query and performance costs. To try the feature out, select Mandatory filter when adding a filter to a Liveboard. For more information about mandatory filters, see Mandatory Liveboard filters.

You can now set your KPI chart to display a percent change comparison between the most recent data point and a custom comparison point. For example, when searching for sales monthly, you can choose to see percent change between this month and the previous month, previous quarter, or previous year. For more information, see Custom comparison points.

You can now use ThoughtSpot Sync to sync data directly from an Answer or custom SQL view to the following apps:

When scheduling a sync with an external app, you can now choose to schedule every 5, 10, 15, 20, 30, and 45 minutes, in addition to the other existing schedule options.

The Parameters feature is now GA. Any user can now create Parameters at the Answer level to optimize their data inquiries. Analysts can also create Worksheet Parameters that are available to all Worksheet users. Use Parameters to run different scenarios with adjustable values, without creating or modifying the constants in formulas for each new value. For example, easily adjust the attribute in an Answer to view your revenue by commit date or order date, as in the following image. To try it out, navigate to any Answer, search, or Worksheet, and select the + button next to the Parameters section in the left panel. For more information, see Creating Parameters and Using Parameters.

You can now see anomalies in time-series KPI charts. Our new anomaly detection algorithm detects data which does not fall within the expected confidence intervals, while allowing for seasonal changes and growth over time. To try it out, open your KPI chart as an Answer, open the chart configuration menu, and select Show anomalies. This feature must be enabled by your administrator. For more information, see Anomaly detection.

KPI change analysis is now GA. You can now use change analysis to directly identify the key drivers for change in your KPIs.

To access this feature, go to your KPI chart and select the caret icon to the right of the percent change label. The Run change analysis pop-up allows you to customize the attribute columns ThoughtSpot analyzes to identify the drivers of change. ThoughtSpot runs change analysis for the most recent data point and the past comparison data point as per the KPI chart’s configuration.

For more information on change analysis, see Comparative analysis. Note that you need SpotIQ privileges to use the KPI change analysis feature.

You can now identify the key change drivers for changes in your metrics in a KPI chart. Select the percent change label or select any two data points on the KPI sparkline to view change analysis on your top 5 columns by usage-based ranking (UBR) and identify the reason behind the observed change. You can view the changes in other columns by clicking Manage attributes and selecting the desired column(s). This feature must be enabled by your administrator.

For more information, see Contextual change analysis.

Apply multiple chart changes at once is now GA. Now users can make multiple configurations and apply them at the same time rather than waiting for each configuration to load separately. For more information, see Reorder labels on the axis or legend.

Optionally disable heatmap charts' data labels, and understand the chart quickly by scanning the heatmap colors. This makes complex heatmap charts with many data values easier to read. To try it out, create a heatmap chart, select the edit chart configuration icon , select Settings, and clear the All labels checkbox. For more information, see Heatmap charts.

We streamlined the file format options when downloading tables and charts from ThoughtSpot. You can download tables in XLSX, CSV, or PDF format, and download charts in XLSX, CSV, or PNG format. To try it out, open a chart or table, and select Download from the more options menu . For more information, see Download a chart or table.

Data labels on a dark background are now GA. Labels that appear on a dark background now display in a lighter text color, making them easier to read. To try it out, navigate to Search data and Create a chart > Change the chart color, and Add labels. For more information, see Reorder labels on the axis or legend.

Arrange Liveboard filters in the filter bar so that they are in a logical, coherent order, by dragging and dropping them to different spots in the list. This makes it easier for viewers of the Liveboard to understand and use the filters on the Liveboard. To try it out, edit any Liveboard, and drag and drop the filters in the filter bar. For more information, see Change the filter order.

We improved the syntax of our text keywords, such as begins with and contains, to provide users with more flexibility. For example, the contains keyword now accepts expressions such as product name contains 'shoes' or 'snickers'. See the Keyword reference.

Additionally, if your query includes two or more contains phrases that modify the same column, ThoughtSpot now combines the phrases using or logic, instead of and. If you have existing Answers or visualizations using this and logic, ThoughtSpot will automatically update your queries after you upgrade to 9.2.0.cl, to ensure they return the same information. See the contains keyword.

This release improves the translated French keywords. To try it out, change your ThoughtSpot locale to French, and include keywords in your search. For more information, see Français (France) keyword reference.

Group aggregate filter enhancements are GA. Previously, when using group_aggregate formulas, users could use query_filters to include filters entered in the search bar in the formula’s value, but it was difficult to write a formula that accepted filters from only a single column.

Starting with this release, you can specify column names within the ‘filters’ section of a group_aggregate formula, and any filter in your search bar referencing that column will be added to the formula. Filters entered in the search bar that are not on columns specified in the third argument of the group_aggregate formula will be ignored. For details, see Group aggregation filters.

To reduce confusion when using aggregate formulas, the level of detail defined in group formulas is now respected at query level. For example, when trying to calculate the contribution of each store’s sales to the entire region, you might use a group_sum formula, where the sales at store level is divided by the sales at regional level. Your search would include the following columns: Revenue, Customer nation, Regional Revenue formula, and Customer region, where Regional Revenue formula = group_sum(Revenue, Customer Region).

According to the old reaggregation behavior, if a customer removed Customer region from the search bar, the level of detail in the formula would no longer be respected and the formula’s denominator would re-aggregate up to the total. In this case, the formula result would display revenue as a percentage of total revenue, rather than as a percentage of regional revenue.

Under the new aggregation behavior, the formula result will continue to display revenue as a percentage of each region, even if you remove Customer region from the search bar. Visualizations will also display the correct aggregation even if the grouping column is not added to the X-axis.

For more information, see Flexible aggregation functions.

ThoughtSpot is updating the Liveboard experience for the mobile app. Mobile app users can now view KPI charts in Liveboards.

When using ThoughtSpot Mobile, you can now press the Load More button at the bottom of a table to see the next 100 rows of data.

When scheduling Liveboards, you can now specify your time zone. To try it out, click on the blue text to the right of the Frequency section and select your timezone from the dropdown menu. For more information, see Schedule a Liveboard job.

In the Search data page, you now see a short description of your selected data source which helps you understand the kinds of answers you can get from that data source.

Data source information includes the name, description, Liveboards created using the data source (if available), tags or classifications (if available), metadata, and experts (top users).

Column information includes the name, description, synonyms (if available), column type, sample values, Parameters, and more.

All of this information saves you time, by helping you find the right data source for the question you want to ask. For more information, see Start a new search with column and data source overview.

In the new Notification preferences section of your user profile, you have the option of receiving email notifications when other users share Liveboards or answers with you, and when they request access to Liveboards or answers which you edit. For more information, see Notification preferences.

There is a ThoughtSpot add-on for Google Sheets:

For details, see ThoughtSpot Google Sheets add-on.

When you create a Worksheet, ThoughtSpot automatically creates synonyms for each column name. Users can easily start searching on your data, without knowing the exact names for every column. For example, if you have a column named Product type, ThoughtSpot might create a synonym such as Product Category. When a user searches for revenue by product category, for example, ThoughtSpot returns sales by product type. To try it out, create and save a new Worksheet. Your column name synonyms appear in the Synonyms column on the Worksheet details page. This feature must be enabled by ThoughtSpot Support. For more information, see Create synonyms for a column.

Cross-Region Disaster Recovery allows ThoughtSpot to recover your ThoughtSpot instance in a secondary cloud region in the case of a failure of the primary cloud region where ThoughtSpot is running. This ensures that if one cloud region fails, ThoughtSpot can start in another region, guaranteeing minimal downtime. For more information, see Data protection and disaster recovery.

Custom geo maps are now GA. Upload your own custom maps and visualize data on geo charts with regions that are specific to you and your organization. To try it out, select Custom maps in the Admin Console. For more information, see Upload custom geo maps.

When creating an alert for a KPI chart, you can now include a custom message in the alert update email. You can use this feature to specify next steps the alert subscribers should take once the KPI reaches a certain threshold.

For data security, you may choose to remove attached files from all scheduled Liveboard emails, so that users need to sign in to view updates in Liveboards they follow. When this feature is enabled, scheduled Liveboard email recipients will receive an email with a link to the Liveboard, which they will need to sign in to see. To ensure that only signed-in users can access your data, contact ThoughtSpot Support to enable this option. For more information, see Scheduled Liveboards.

You can now create connections from ThoughtSpot to SQL Server. For details, see SQL Server.

You can now create connections from ThoughtSpot to Generic JDBC databases. For details, see Generic JDBC. To enable this feature, contact ThoughtSpot Support.

To enable this feature, contact ThoughtSpot Support.

You can now connect to and query Amazon Aurora and Amazon RDS for PostgreSQL databases.

When creating a sync to HubSpot, you can now use email as a mapping field on the Contacts object type, as well as ContactID.

When creating a connection to Databricks, you can now enter your catalog name under the Catalog field. For more information, see Add a Databricks connection.

By default, scheduled and downloaded Liveboard PDFs now render at a width of 1920 pixels. This ensures that most Liveboard PDFs you download or schedule look the same as the Liveboards you see in ThoughtSpot. You can also change this default width to a different width for your company, by contacting ThoughtSpot Support. To try the new default width out, download a PDF of a Liveboard. For more information, see Download a Liveboard as a PDF and Schedule a Liveboard job.

Use the Object Usage Liveboard to easily track how specific objects (Liveboards, Answers, Worksheets, or tables) are trending over time. Our new visualizations allow you to understand adoption on the object level, and filter on a specific user or users to see usage on the user level.

For more information, see Object Usage Liveboard.

TML for Monitor alerts is now GA. Export, edit, and import KPI Monitor alerts programmatically, using ThoughtSpot Modeling Language. To try it out, export a Liveboard that contains Monitor alerts and its associated objects. For more information, see TML for Monitor alerts and Monitor KPIs in your data.

Use TML to directly edit a join condition by editing the TML file’s on parameter. This simplifies the previous process of deleting and recreating the join in TML or the ThoughtSpot application. To try it out, edit a TML file, and update a join’s on parameter. For more information, see TML join syntax.

Remove columns from tables in ThoughtSpot by deleting them in the tables' TML files. To try it out, edit a table TML file, and delete a column’s name and properties. We support deleting a column from a TML file and removing that column’s dependencies in the same zip file import. Note that if the deletion of a column fails, the whole import will fail. For more information, see Deleting columns, joins, and RLS rules.

We standardized the method of creating joins for all data objects in ThoughtSpot. When creating a join from a view, users are now directed to a Create join page where they can choose a connection, name the join, and map the source and destination columns.

Adjust Parameter values at runtime from Liveboard or Answer URLs, or using REST API v1 requests. To try it out, add the Parameter runtime override syntax to any Liveboard or Answer with an active Parameter. For more information, see Runtime overrides for Parameters and Runtime Parameters.

For any column with a high cardinality, which is defined as more than 100,000 unique column values, ThoughtSpot does not index the column if the index type is DEFAULT. This means the column values do not show up in search suggestions. Previously, this limit was 1 million unique column values. This change improves performance, and reduces index times and memory usage.

If you have existing columns with a cardinality between 100,000 and 1 million, and you use these columns' search suggestions regularly, ThoughtSpot will continue to index these columns. For more information, see High cardinality and performance.

In this release, we introduce Private Preview features. Private Preview features are complete, and available for select customers to try before they reach GA, when they are enabled by default. To try them out, contact ThoughtSpot Support to enable these features. Then, administrators can selectively enable or disable this feature set for users and groups, using the Can preview ThoughtSpot Sage user privilege. For more information, see Private Preview features and Understanding privileges. Private Preview features are disabled by default.

For new features and enhancements introduced in this release of ThoughtSpot Everywhere, see ThoughtSpot Developer Documentation.

June 2023 9.3.0.cl

In this release, ThoughtSpot Sage features are available for administrators to enable, without contacting ThoughtSpot Support. ThoughtSpot Sage makes it easier for you to get answers and insights from your complex data. ThoughtSpot Sage uses a combination of GPT-3, GPT-3.5T and GPT-4, and the large language models (LLM) of the Microsoft Azure OpenAI Service.

Sage features include:

Once enabled, your ThoughtSpot administrator can grant access to the Natural language search and AI-suggested searches features for users in a group by assigning the Can preview ThoughtSpot Sage user privilege to the group. You do not need this privilege to access AI-generated Worksheet column synonyms.

You can now search using natural language to access AI-generated answers to your business questions. This feature is now available for administrators to enable, without contacting ThoughtSpot Support.

The AI-suggested searches feature is now available for administrators to enable, without contacting ThoughtSpot Support. This feature helps you think of relevant questions to ask of your own data source, by providing examples of searches. This feature is designed to help you get started if you are a new user of Search Data, or if you are working with a new Worksheet. This feature must be enabled by your administrator.

ThoughtSpot Cloud will be available on Google Cloud Platform in addition to AWS. Customers can choose whether they want their SaaS deployed on AWS or GCP.

ThoughtSpot on GCP can be purchased directly from ThoughtSpot but will also be available to purchase in the Google Cloud Marketplace, the BigQuery Partner Center, and Google Workspace Marketplace. Customers can also use their committed spend and credits on Google Marketplace to purchase ThoughtSpot. To migrate to GCP, contact ThoughtSpot Support.

Liveboard cross filters are now GA. Right-click any data point in your Liveboard and easily filter the entire Liveboard by that value. For example, if you right-click California in a geo chart on your Liveboard, and then select Filter, all the other visualizations only show data from California. A visualization for Total sales by city, for example, would only show sales for cities in California. This feature is also called brushing and linking.

The Liveboard note tiles feature is now GA. Add notes to your Liveboards, so that viewers of your Liveboards can quickly understand how to use the data and visualizations you gathered. Note tiles can provide an overview of a Liveboard or visualization, company branding through images, definitions of terms used in the Liveboard, and so on.

You can now include attributes in pivot table cells, as well as measures. This is useful for cases when you want to summarize text data such as user role, pass/fail, or other attributes.

Arrange Answer filters in the filter bar so that they are in a logical, coherent order, by dragging and dropping them to different spots in the list. This makes it easier for viewers of the Answer to understand and use the filters on the Answer.

To enable business users to answer questions related to mix, contribution, share of, percentage, and total ratios, we added the keywords “of” and “percentage of”. These keywords help answer questions such as “What is the share of a certain store’s sales as a percentage of the regional whole?” Previously, searches such as these required the use of a group aggregate formula, which presented a barrier to users.

When you scroll down in a Liveboard, the Liveboard name, menu options, and filter and tab bars remain visible. You don’t need to scroll back up to the top to edit the Liveboard or check if you’re filtering by a specific store. You can see the filters or parameters applied to the Liveboard, select the Edit button or more options menu, and view which tab you’re on from any spot in the Liveboard.

You can now configure the legend position of a chart, or hide the legend. The available options are left, right, above, and below the visualization. To try it out, open the chart configuration menu, select Settings, and choose between Legend on right, Legend on left, Legend on top, and Legend on bottom. You can select the Show legend toggle to hide or show the legend.

To reduce confusion when using aggregate formulas, the level of detail defined in group formulas is now respected at query level. For example, when trying to calculate the contribution of each store’s sales to the entire region, you might use a group_sum formula, where the sales at store level is divided by the sales at regional level. Your search would include the following columns: Revenue, Customer nation, Regional Revenue formula, and Customer region, where Regional Revenue formula = group_sum(Revenue, Customer Region).

According to the old reaggregation behavior, if a customer removed Customer region from the search bar, the level of detail in the formula would no longer be respected and the formula’s denominator would re-aggregate up to the total. In this case, the formula result would display revenue as a percentage of total revenue, rather than as a percentage of regional revenue.

Under the new aggregation behavior, the formula result will continue to display revenue as a percentage of each region, even if you remove Customer region from the search bar. Visualizations will also display the correct aggregation even if the grouping column is not added to the X-axis.

You can now identify the key change drivers for changes in your metrics in a KPI chart.This feature is now available for administrators to enable, without contacting ThoughtSpot Support. Select the percent change label or select any two data points on the KPI sparkline to view change analysis on your top 5 columns by usage-based ranking (UBR) and identify the reason behind the observed change. You can view the changes in other columns by clicking Manage attributes and selecting the desired column(s).

When you create a Worksheet, ThoughtSpot automatically creates synonyms for each column name. This feature is now available for administrators to enable, without contacting ThoughtSpot Support. Users can easily start searching on your data, without knowing the exact names for every column. For example, if you have a column named Product type, ThoughtSpot might create a synonym such as Product Category. When a user searches for revenue by product category, for example, ThoughtSpot returns sales by product type.

This release includes the following improvements to the dbt integration:

You can now create connections from ThoughtSpot to Amazon Aurora PostgreSQL.

You can now create connections from ThoughtSpot to Amazon RDS PostgreSQL.

You can now create connections from ThoughtSpot to MySQL databases.

Previously, when multiple database connections were configured to receive CSV uploads without a specified table or worksheet for collocation, the files were randomly uploaded to one of the connections. With this new enhancement, administrators and data managers now have the ability to specify a default target connection for such uploads.

To set the default target connection, navigate to Data > Connections and use the new CSV Upload Default toggle. By specifying a default connection, you ensure that CSV files without a designated destination will be uploaded to the chosen connection. This enhancement provides better control and organization of CSV uploads, making it easier to manage data in scenarios where end-users do not select a collocation for their uploaded files.

Admin users will be able to access a Liveboard reflecting the pricing model purchased and associated with their cluster. Query-based pricing customers see the Billable Query Stats Liveboard, while time-based pricing customers see the Credit Usage Liveboard.

For new features and enhancements introduced in this release of ThoughtSpot Everywhere, see ThoughtSpot Developer Documentation.