What’s New in v17.2.16 for Self-Service & Embedded Analytics

TL;DR: Bold BI® v17.2.16, released on November 10, 2026, adds custom AI prompts for individual widgets, dashboard and data source tags, Snowflake Semantic Layer support, PowerPoint export templates, and general availability of Bold Data Hub. Embedded applications must update to the v17.2.16 Embed SDK.

Introduction

We’re excited to announce the release of Bold BI v17.2. This release focuses on three things: giving teams more control over AI-generated insights, making large analytics deployments easier to organize, and simplifying how data is modeled and moved.

Bold BI v17.2.16 brings widget-level AI prompting, AI support for dashboard parameters and multi-tab dashboards, tag management for dashboards and data sources, Snowflake Semantic Layer support, and new financial expressions. Bold Data Hub is now generally available, with pipeline export and import, reusable connection credentials, and faster SQL Server refresh options.

Whether you build dashboards for internal teams or embed analytics in your own product, here’s what’s new in Bold BI v17.2.

AI enhancements

This release gives administrators more control over what AI summaries say, and extends AI to more dashboard contexts.

Customize AI summaries for individual widgets

A sales widget and an operations widget on the same dashboard often need different kinds of explanation. Until now, custom prompts applied only to dashboard summaries.

Bold BI v17.2.16 extends custom prompting to individual widgets. Authors can write instructions for each widget, such as “focus on revenue variance against target,” and the AI-generated widget summary follows them.

Use dashboard parameters and multi-tab dashboards with Bold BI AI

The AI Assistant now supports dashboard parameters, so questions are answered using the parameter values currently applied to the dashboard. The AI Agent now supports multi-tab dashboards, so users can work with AI across every tab of a dashboard, not just a single view.

Designer enhancements

The dashboard designer gains two new features and a set of formatting, export, and layout improvements.

Switch fields and date ranges in view mode

Report viewers often need to compare a different measure or period without asking the dashboard author for a new version.

Bold BI v17.2.16 adds dynamic field selection and dynamic date range selection with offset in view mode. A viewer can, for example, switch a chart from revenue to margin, or compare the last 30 days with the same period offset by one year.

Export dashboards with PowerPoint templates

Exported slides often need reformatting before they reach an executive review.

Bold BI already exports dashboards to PowerPoint. In v17.2.16, PowerPoint export with templates lets you export dashboards and widgets into predefined templates that carry your layout and branding.

Why it matters: Exports are ready to present without manual cleanup.

Format, color, and export dashboards with more precision

These improvements give authors finer control over how numbers, colors, and exports look:

• Data source-level measure formatting: Set number formats once in the data source, and they apply across every widget and dashboard that uses it.

• Dashboard-level date format: One setting controls date display across supported widgets.

• Dynamic color mapping: Classic and Modern charts can take colors from a data field containing HEX codes, RGB values, or supported color names.

• PDF export: Standard paper sizes are improved, and a new fit-to-page option exports the whole dashboard on one page.

• Page-wise chart PDF export: Available for Bar, Column, Stacked Bar, and Stacked Column charts without Row bindings.

• Data label rotation: Rotate data labels in Pie and Doughnut charts.

• Category-based paging: Classic and Modern charts page by category, so each category is shown in full.

• Grid layout: Fixed Height and Fit to Content options remove unused space in Classic and Modern Grid widgets, and new vertical alignment settings control cell content position.

• Indian number formats: Auto Measure Formatting now supports Thousand Crore and Lakh Crore.

Navigate, migrate, and load dashboards faster

• Pivot Grid linking to dashboard tabs: Pivot Grid widget linking can open a specific tab in the target dashboard.
• Classic to Modern widget migration API: On-premises deployments can migrate Classic widgets to Modern widgets through an API.
• Mobile layout: Widgets inside Tab and Combine widgets now stack vertically on mobile screens.
• Loading performance: Optimized dashboard and widget JavaScript and CSS resources reduce dashboard load time.

Why it matters: Authors spend less time reworking layouts, and viewers get consistent dashboards on desktop and mobile.

Server and administration enhancements

As deployments grow to hundreds of dashboards, finding content and controlling what each user sees becomes harder.

Organize dashboards and data sources with tags

Bold BI v17.2.16 adds tag management for dashboards and data sources. Administrators and users can add, manage, and remove tags in the user interface or through APIs. Tags appear on listing pages, and you can filter dashboards and data sources by tag, for example by department, customer, or project.

Control what each user role sees

• Role-based visibility for custom UI: In the custom UI configuration file, administrators can specify which menus and actions Admin and Non-Admin users see within a tenant. If no roles are specified, the existing visibility behavior is kept.
• Hide profile menu items: Configuration settings can hide profile and sub-menu items for non-admin users.
• API key retrieval: A new API lets users retrieve their own API key details for automation.
• Choose Columns: The Data Sources listing page lets each user show or hide grid columns.

Why it matters: Administrators can give each audience a simpler interface without custom code.

Data Sources highlights

This release makes data modeling faster for analysts and adds new calculation options.

Use Snowflake semantic models in Bold BI

Many organizations define business metrics once in Snowflake to keep them consistent across tools.

Bold BI v17.2.16 adds Semantic Layer support for the Snowflake connector, so dashboards can use Snowflake semantic models directly.

Why it matters: Dashboards built in Bold BI use the same governed metric definitions as the rest of your data stack.

Build data models faster with automatic relationship detection

When tables are added to the Query Canvas, Bold BI now detects relationships automatically using primary and foreign keys, or matching column names with compatible data types.

• Auto-load related tables: Related tables load automatically based on detected relationships.
• Multi-table selection: Add several tables at once.
• Key indicators: For Live connections, the Data Model Designer shows primary key and foreign key icons beside column names, including composite keys.

Model complex scenarios with less effort

• Role-playing dimensions: One dimension table, such as Date, can serve several roles (Order Date, Ship Date, Delivery Date) without duplicating it.
• Relationship-aware aggregation: Measures from related tables at different grains are aggregated correctly, without row multiplication, for SUM, COUNT, AVG, MIN, MAX, and DISTINCTCOUNT.
• Bulk display name updates: The Data Source Column API’s new UpdateColumnDisplayName action renames columns in bulk and carries the change to dependent data sources and dashboards, including filters, parameters, expressions, and widget bindings.
• Total column count: The Data Source page shows the total column count, including expression columns.

Calculate returns and period comparisons in expressions

• XIRR and IRR: Calculate internal rate of return directly in dashboard expressions for financial and investment analysis.

• PREVIOUSPERIOD: Build period-over-period comparisons at the data source level in the PostgreSQL and SQL Server connectors.

• NCHAR: Generate Unicode characters in String expressions for symbols and multilingual text.

Connect to SQL Server with service principal authentication

The SQL Server connector now supports service principal authentication through Microsoft Entra ID (Azure AD). Connections no longer depend on an individual user’s account, which simplifies credential management and automated access.

Why it matters: Security teams can manage database access centrally.

Bold Data Hub highlights

Bold Data Hub, the ETL component inside Bold BI, reaches a major milestone in this release.

Bold Data Hub is now generally available

The Beta label has been removed from Bold Data Hub. It now offers 48 connectors for moving data from many sources into one central data store.

Move pipelines between environments and reuse credentials

• Pipeline export and import: Move pipeline configurations between development, testing, and production without recreating them.

• Connection-level custom attributes: Store credentials once and reuse them across multiple data source connections.

Refresh SQL Server data more efficiently

Incremental refresh is already available for SQL connectors in Bold Data Hub. Bold BI v17.2.16 adds two new methods for SQL Server:

• Primary key-based incremental refresh: Identifies newly added records by the primary key column, avoiding full reloads.

• SQL Server Change Tracking: Incremental refresh now also picks up updated and deleted records, not just new ones.

Connect and automate more pipeline tasks

• Salesforce OAuth: Connect to Salesforce through an OAuth authorization flow instead of entering account credentials.

• Transform table APIs: Create, update, and delete transform tables through REST APIs.

• Destination data types: View and change column data types in the Destination tab before loading.

Embedding highlights

The updates below add commenting, custom loading screen branding, and tag-based filtering to dashboards in applications built with embedded analytics.

Add dashboard comments to SDK-embedded dashboards

Dashboard comments now work in SDK-based embedding, so your customers can discuss insights without leaving your application.

Customize the loading screen logo in SDK-embedded dashboards

You can replace the loading screen logo in SDK-embedded dashboards with your own, so the embedded experience matches your product’s branding from the first second.

Filter embedded content with tags

Full Server Embedding can now filter dashboards by tag, using direct tag values or attribute-based mappings. Each user sees only the dashboards relevant to them, without duplicating content.

Other bug fixes and minor improvements

Bold BI v17.2.16 also includes fixes across the Designer, Server, Data Sources, Data Hub, Embedding, IDP, and AI modules.

• Reliability: Fixed issues in Pivot Grid COUNTD and AVG expressions, manual sorting after auto-refresh, filter retention in tab-mode dashboard linking and Excel export, and expression configurations after dashboard upload.
• Compatibility: Improved behavior with Google BigQuery, Databricks on Azure Linux, PostgreSQL, Oracle, Salesforce, Asana, and custom widgets when Content Security Policy is enabled.
• Authentication: Fixed issues in JWT session logout, Keycloak group mapping, and OpenID Connect logout.
• Security: Upgraded PuppeteerSharp to v25.15.0 with the matching Chrome version, and applied HTML sanitization to image and PDF exports.
• Usability: Standardized dialog button placement and improved validation messages, context menu positioning, and license error messages.

Conclusion

Bold BI v17.2.16 gives teams more control over AI insights, makes large deployments easier to organize, and simplifies data modeling and pipeline management, with Bold Data Hub now generally available.

Ready to try it? Download Bold BI v17.2.16 or start a free trial to explore the latest self-service analytics features. For the complete list of features, improvements, and fixes, see the v17.2.16 release history. If you have questions, contact our support team.