Skip to main content

​
AI recommendations for validators

Configuration for validator recommendations with LLM credentials

Configuration for validator recommendations with LLM credentials

Automatically generate tailored validators for your data sources with AI Recommendations. Instead of configuring validators one by one, you can now profile your data and receive a curated list of suggested validators, windows, filters, and segmentations to apply all at once. Two modes are available:
  • AI recommendations: LLM-powered, context-aware suggestions (requires LLM credentials)
  • ML recommendations: Statistical machine learning and heuristics-based suggestions
AI recommendations are currently available only for tabulated data sources. Support for Custom SQL sources and views will be added in a future release.
To learn more, see Generating Validator Recommendations in our product documentation.

​
Referential integrity validator

Fields mapping for referential integrity validator

Fields mapping for referential integrity validator

A new validator type is now available for validating record-level consistency across two datasets in your data warehouse. The Referential Integrity Validator verifies that keys—and optionally values—in a primary source have corresponding matches in a reference source, helping you identify orphaned records or data mismatches between related tables. Key capabilities include:
  • Key field mapping: Join records between sources using one or more key fields (such as, order_id to transaction_id)
  • Value field comparison: Optionally verify that matched records have identical values in specified fields
  • Flexible windowing: Configure independent windows for primary and reference sources (Global or Tumbling)
  • Filtering and segmentation: Apply filters and segmentations to narrow the scope of checks
For configuration details, see Referential Integrity in our product documentation.

​
ServiceNow integration

Configuration for ServiceNow integration

Configuration for ServiceNow integration

Validio now integrates with ServiceNow for streamlined incident management. Data quality issues, source errors, schema changes, and segmentation limit errors detected in Validio can automatically create and update incidents in ServiceNow. Key capabilities include:
  • Automatic incident creation: Validio incidents automatically create corresponding ServiceNow incidents, grouped by source
  • Bidirectional syncing: Status updates, owner assignments, and comments sync between both systems
  • Customizable routing: Use notification rules to filter which incidents are tracked in ServiceNow
For setup instructions, see ServiceNow in our product documentation.

​
Amazon Bedrock integration

Configuration for Amazon Bedrock credential

Configuration for Amazon Bedrock credential

You can now configure Amazon Bedrock as an LLM credential in Validio. This enables you to use AI foundation models from Anthropic, Meta, Mistral, and Amazon for SQL query generation and validator recommendations. With an Amazon Bedrock credential, you can:
  • Generate SQL queries: Use AI assistance when writing custom validators and SQL filters
  • Get validator recommendations: Leverage LLM-powered data profiling to suggest pre-configured validators for your sources
For setup instructions, see Amazon Bedrock in our product documentation.
​
9 February 2026

​
Fixed issues

  • Fixed issues with concurrent MCP sessions (UI-5824)
​
6 February 2026

​
Fixed issues

  • Fixed a bug where creating or updating a source would inadvertently delete descriptions in the schema fields (VR-5144)
  • Improved user activity tracking when applying recommendations (UI-5817)
​
6 February 2026

​
Fixed issues

  • Updated the profiling API to use the correct schema (UI-5804)
​
5 February 2026

​
Deprecations and removals

  • Type-specific Entity APIs: The dedicated create and update endpoints for specific entity types are now deprecated and scheduled for removal in a future release. To facilitate the transition, @internal restrictions have been removed from the following mutations, making them publicly accessible: (UI-5648)
    • sourceCreate, sourceUpdate
    • validatorCreate, validatorUpdate
    • filterCreate, filterUpdate
    • windowCreate, windowUpdate
    • credentialCreate, credentialUpdate
    • channelCreate, channelUpdate

​
Fixed issues

  • Fixed an issue with placeholder text in “Convert to source” wizard (UI-5709)
  • Fixed an issue to ensure recommended validators initialize with backfill (UI-5773)
  • Fixed a pagination issue on Sources page (UI-5763)
  • Fixed a form validation issue when uploading Kafka credentials (UI-5748)
  • Added dialect-specific SQL examples for SQL validator agents (UI-5778)

​
Features and updates

  • Added query logs support for Databricks (VR-4793)
  • Added support for Amazon Bedrock credential (UI-5755, UI-5769, UI-5771)
  • Added support for ServiceNow integration (UI-5661)
  • Added new validator recommendation agent (UI-5712, UI-5768, UI-5766, UI-5782, UI-5780)
  • Added new referential integrity validator with RCA support (UI-5634, UI-5684)
  • Added support for including comments in SQL queries for Debug (VR-5052)
  • Implemented a new trigger logic to handle high-frequency level shifts. The system now automatically initiates a reverting level shift when a cluster of multiple shifts occurs near a recent incident group (AL-643)
  • Implemented a change to merge direct and team roles in Namespaces (UI-5761)