v3.0

This release introduces the new Clarity UI design system, enhanced DQ-AI Assistant capabilities with cross-table awareness, and natural language schema generation for the PDF connector. It also delivers faster performance through virtualization and prefetching, improved permission assignment during static file loading and virtual table creation, and more control with column-level scoping at the ruleset level.

New Features

Clarity UI – New design system

• Introduces a unified and modern design system across the DQC Platform

• Improves usability with clearer layouts, better visual hierarchy, and more consistent components

• Enhances readability through better contrast on components and richer colors

• Keeps existing structure and navigation unchanged, ensuring no disruption to established user flows

Improved cross-table awareness in DQ-AI Assistant

• Enhances the DQ-AI Assistant's understanding of relationships across tables to improve answer quality and rule suggestions

• SAP Semantic Relationship Tool

  • Enriches the assistant with knowledge of SAP table relationships

  • Improves recognition of SAP tables and how commonly used tables are connected

  • Enables more context-aware responses and improved suggestions for e.g., cross-table consistency checks

• Access to connected tables

  • Allows the DQ-AI Assistant to access tables within the same connector

  • Improves answers and rule suggestions by leveraging broader data context across tables

In the screenshot below, the user asks: "Make sure we don’t have product variants that belong to non-existing products." The Assistant responds by generating a cross-table rule that enforces this condition.

PDF schema generation

• Introduces a prompt field in the PDF connector menu where users can describe in natural language which information should be extracted from PDF files

• Automatically generates a schema based on the input and provides a preview for validation

• Removes the need to manually create YAML schemas, making the PDF connector much easier and faster to use

Performance improvements

• Fast loading even for very large tables

  • Large tables are now loaded using virtualization, significantly improving performance

  • Enables smooth scrolling and interaction even for very large datasets

• Prefetching of common user flows

  • Introduced prefetching to reduce load times across the DQC Platform

  • Frequently used elements will now be preloaded (e.g. favorite tables when entering the table overview)

  • Enables faster navigation and a more responsive user experience

Permissions during creation of static files and virtual tables

• Users can now assign permissions directly while loading static files or creating virtual tables

• Eliminates the need to configure access rights only after creation, improving workflow efficiency

• Permissions set during creation can still be adjusted later at the table level

• Enhances data privacy and security while providing a smoother and more intuitive user experience

Column filter for ruleset scope

• Define exactly which columns are included in a ruleset, improving control and performance

• Limits visible data to user-selected columns to support data privacy and security requirements

• The column filtered ruleset scope is applied consistently across samples (issues and valid samples), downloads, profiling, and filters in rules, alerts, missions, and elsewhere

Small improvements and bug fixes

• Rule import now also validates rule metadata

• Improved static connector robustness and large file handling

• Fixed incorrect sorting by status in the rule repository

• Fixed an edge case where the duplicate survivor node was not returning improvements for all rows

Breaking Changes

• None in this release