Skip to content

Changelog – Release 2026.06

With Release 2026.06, digna takes a major step forward in automation, extensibility, and platform usability.
This release introduces the new digna Python SDK, official Docker deployment support, a refreshed dashboard experience, and enhanced portability for validation rule management.

Watch the Release

What's New in digna | The Major Release of 2026 — a walkthrough of this release on the digna YouTube channel.

New Features

digna Python SDK – Automate Everything with Python

  • Install via: 
    pip install digna-sdk
  • Programmatically manage and automate digna using Python
  • Create and configure projects via code
  • Trigger inspections and monitoring executions
  • Manage datasets, rules, and configurations programmatically
  • Profile tables and extract metadata insights
  • Export profiling and data quality results to external repositories and systems
  • Integrate with notebooks, orchestration tools, and CI/CD pipelines

Impact: Enables full infrastructure-as-code and deep automation of data quality and observability workflows using Python.

Docker Support – Simplified Deployment & Operations

  • Official Docker image support for digna
  • Fast and consistent setup across environments
  • Simplified onboarding for development, test, and production
  • Easy integration with Kubernetes and container platforms
  • Improved portability and reproducibility of deployments

Impact: Makes digna easier to deploy and operate in modern cloud-native architectures.

QueryMode – Flexible SQL Execution Strategy

Configure query execution strategy: Single or Combined mode

Single Mode: Each statistic is calculated with one dedicated SQL query

  • Ideal for large datasources where memory constraints are a concern
  • Prevents combined query resource exhaustion (out of memory, spool limits)
  • Higher query count but lower per-query memory footprint

Combined Mode: All statistics are computed within a single SQL query

  • Reduces total query count and network overhead
  • Optimizes performance when datasources are manageable in memory
  • More efficient for frequent, parallel executions

Impact: Gives users fine-grained control over query execution to balance performance, resource usage, and memory safety based on their datasource characteristics.

Configurable Prediction Model

The model behind anomaly detection now weighs competing explanations of each series — a pattern combined with a reading of the most recent observations — and blends their forecasts by how strongly each is supported. A single extreme value can no longer leak into the following predictions.

Two settings on the data source's new Model tab steer it, each from 0.0 to 1.0 with 0.5 as the default:

  • Break Sensitivity – how quickly a jump to a new level or a turning trend is believed rather than treated as outliers
  • Model Complexity – how much structure the model looks for, from calendar effects and a single level shift up to unknown cycles, month-day effects and monthly resets

Both can be restored to their defaults at any time. See Model Settings for how each one acts.

Impact: Gives users control over the prediction model itself, alongside the existing Sensitivity and Memory settings on the tolerance band, now on the Thresholds tab.

Anomaly Notification Controls

  • New Notifications tab on the data source's anomaly settings:
  • Minimum Alerts – how many failed checks (not uncertain ones) an inspection needs before a notification is sent (default 1)
  • Pause After Notification (Days) – how long a subscription stays silent after notifying about the data source (default 0, no pause)

Redesigned Dashboard Experience

  • Modernized and improved UI/UX design
  • Clearer navigation and structure
  • Better visibility of monitoring results and data quality insights
  • Improved readability of alerts, statistics, and dashboards
  • Faster access to key operational information

Impact: Improves usability and daily productivity for all users.

Extended Import & Export for Validation Rules

  • Enhanced import/export functionality for validation rules
  • Easier migration between environments and projects
  • Improved reuse of standardized rule sets
  • Better governance and rule lifecycle management
  • Simplified collaboration across teams

Impact: Enables scalable and consistent data quality governance across the organization.

Platform Enhancements

  • Full Python SDK integration for automation
  • Containerized deployment via Docker
  • Improved UX through redesigned dashboard
  • Expanded portability of validation logic

Who Benefits from This Release

  • Data Engineers: automation, SDK usage, pipeline integration
  • Platform Teams: simplified deployment via Docker
  • Data Governance Teams: reusable validation rule management
  • Analytics Teams: improved usability and insights visibility

CLI Updates

  • Added SDK integration support
  • Improved import/export workflows
  • General stability and performance improvements