Bridge the data health gap between technical and business assets

Share on:

Data complexity is exploding, leaving manual tracking in the dust. According to KPMG, 92% of data leaders admit that high-quality data products are their top priority, yet Gartner reports that most implementations fail due to poorly defined data product standards. The gap? Visibility. You can’t scale what you can’t see, and you can’t lead with data you don’t trust.

What’s new: Data quality tab enhancements

Collibra is proud to announce a major overhaul of the data quality tab, an evolution in how quality metrics are visualized and managed across the enterprise. This feature introduces the ability to leverage derived relations in aggregation paths, allowing users to propagate data quality scores across complex data ecosystems. By leveraging the power of our metadata graph, the quality tab allows users to roll up scores from physical columns all the way to high-level business assets and data products.

This feature ensures that data quality is no longer siloed at the technical level. Instead, it is aggregated into meaningful scores that reflect the health of an entire data product or business domain. This enhancement introduces granular controls for score frequency, dimension assignments and derived relations support.

How the quality tab enhancements help

The modern data landscape is often too fragmented for traditional linear tracking. Organizations struggle to bridge the gap between technical assets and business context, leading to inaccurate quality reporting and wasted steward time. With the enhanced quality tab you can solve these challenges by automating the roll-up of quality metrics. This ensures that business leaders see the real-time health of their data products without manual intervention.

With the enhancements in the quality tab, you can:

Ensure score relevancy: Match reporting frequency to your data’s unique lifecycle supporting monthly, weekly or daily refresh rates
Support complex data structures: Handle "branching" logic for data products that draw from multiple schemas, tables or datasets
Enhance precision: Improve quality reporting with configurable filters for monitor types and dimensions for any asset like business rules, data products and policies
Reduce “time-to-trust”: Make data quality scores visible everywhere, from asset lists to search results in the data marketplace
Eliminate manual effort: Automate the tedious task of rolling up DQ scores across thousands of assets

How the quality tab enhancements work

The new quality tab functions by utilizing the Collibra metadata graph, powered by Neo4j. Unlike the previous version which relied on standard PostgreSQL queries, this new version is designed to execute complex, multi-hop graph queries that can traverse many relations without degrading performance. The system follows a defined "aggregation path"—a sequence of connected relations—to identify all physical columns that contribute to a business asset's overall health.

When a Data Quality (DQ) job runs, the resulting scores are aggregated along these graph paths. For data products, this means the system can simultaneously look at multiple "branches"—such as a Data Product Port linking to a Schema, then a Table and finally individual Columns—and consolidate those disparate scores into a single metric. Integration with the broader Collibra Platform is seamless.

The aggregated scores are stored as attributes on the assets, making them available for filtering in the Data Marketplace and for visualization in Control Tower in May. This allows the platform to act as a live control plane, where a change in a single column's quality score can instantly trigger a re-calculation of the health status for every linked business report or AI model.

Why you should be excited

The enhancements to the quality tab fundamentally change how data quality is measured and trusted across your organization. By automating the roll-up of quality scores from technical assets to high-level data products, Collibra delivers unified, real-time visibility that empowers every stakeholder to act with data confidence. Here is how this innovation will specifically benefit key roles and use cases across your business:

Data Stewards: Take advantage of data quality scoring in critical data elements and policies to validate compliance across all asset types without manual tracking, freeing your time for strategic governance tasks.
Governance Managers: Use the unified DQ score to instantly judge the quality of any data asset in your portfolio and identify high-risk areas before they impact the business
Data Scientists: Save time on data prep by discovering best-fit, certified datasets with proper quality scores.
CIOs & CTOs: Gain a centralized view of the entire AI and data landscape, allowing you to make confident investment decisions based on technical and business context

This feature helps these users with:

Data product certification: Automatically aggregate quality scores from dozens of disparate tables into a single certification status for a consumer-facing data product
Regulatory lineage for finance: Trace the quality of "Critical Data Elements" (CDEs) by rolling up column-level DQ results through the semantic layer to final regulatory reports
AI model training safety: Prevent the use of unvetted data in AI models by creating a path that links model deployments to foundational model policies and the quality of underlying training datasets

Key takeaways about quality anywhere

This launch tightly integrates Data Governance and Data Quality into a single, automated experience. By integrating this new capability with the rest of the Collibra innovations ensures that quality is tracked across critical business assets. Here are the three things you should not forget about quality anywhere:

Join Collibra’s Spring Product Premiere to learn how this new features provides:

Automation at scale: Move from manual, labor-intensive checks to a reliable, recurring system that determines data readiness status automatically
Unified visibility: Provide a single pane of glass for data and AI health, regardless of where the data is stored or computed
Trust through traceability: Ensure that every data product and AI model is grounded in dependable, high-quality data with end-to-end lineage