26.2.0
Test Data Management introduces new features and enhancements focused on improving data generation, automation, and usability.
Software Version: 26.2.0
Publication Date: April 2026
New Features¶
-
Added five new predefined list generators:
- Payment Methods
- Industry Sectors
- Education Levels
- Interaction Status
- Membership Tiers
-
Expanded the Regex Generator with ten new patterns:
- MAC address
- SWIFT/BIC bank code
- JWT (JSON Web Token)
- ISBN-13
- IATA airport code
- Flight number
- Semantic versioning
- MIME type
- Vehicle Identification Number (VIN)
- IMEI number
-
Enhanced flow generation by analyzing table relationships, automatically splitting them into multiple flows and assigning them to a newly created project.
-
Introduced AI-driven sample data generation based on user prompts, enabling rapid creation of relevant datasets for testing and demonstration.
-
Added automated data population by analyzing sample datasets, inferring column structures, and generating synthetic data aligned with the original format.
-
Introduced AI-based data generation capabilities that learn from production data and generate synthetic datasets while preserving statistical characteristics.
-
Introduced Shift Flow, enabling data export to file formats such as JSON, CSV, and Parquet.
Enhancements¶
-
Improved the Dashboard interface with increased detail and enhanced functionality.
-
Enhanced support for special characters in table and schema names.
-
Improved AI capabilities for generating context-aware data lists (e.g., names, addresses, product codes) tailored to specific scenarios.
-
Updated the Script Flow package version list.
Critical Updates for Script Flow¶
Breaking Changes
This release includes major library upgrades that may impact existing scripts. Code updates and regression testing are strongly recommended.
Package Migrations & Replacements¶
Removed Packages¶
- kafka-python (2.0.2)
- pylint (3.3.5)
- mysql-connector-python (>=8.0.0)
- teradatasqlalchemy (>=20.0.0.0)
- sqlalchemy-databricks (~0.1)
- sqlalchemy-redshift (0.8.12)
- sqlalchemy-vertica (0.0.5)
- clickhouse_sqlalchemy (0.2.6)
Added Packages¶
- confluent-kafka (2.13.2)
- ruff (0.15.5)
- PyMySQL (1.1.2)
- databricks-sql-connector (4.2.5)
- databricks-sqlalchemy (2.0.9)
- redshift-connector (2.1.12)
- pandas (2.2.3)
- SQLAlchemy (2.0.48)
- sqlalchemy-vertica-python (0.6.3)
- clickhouse-connect (0.13.0)
- Authlib (1.6.9)
- azure-core (1.38.2)
- google-cloud-aiplatform (1.140.0)
- Werkzeug (3.1.6)
- filelock (3.25.0)
- pyasn1 (0.6.2)
- python-multipart (0.0.22)
- wheel (0.46.3)
- aiohttp (3.13.3)
Updated Packages¶
| Package | Old Version | New Version |
|---|---|---|
| pyspark | 3.3.2 | 4.1.1 |
| numpy | <2 | 2.4.2 |
| pyarrow | ~16.0.0 | 23.0.1 |
| apache-airflow | 3.1.3 | 3.1.7 |
| scikit-learn | ~1.0 | 1.8.0 |
| urllib3 | 1.26.20 | 2.6.3 |
| pydantic | ~2.0 | 2.12.5 |
| boto3 / botocore | 1.35.23 | 1.42.61 |
| azure-identity | ~1.12 | 1.25.2 |
| azure-storage-blob | ~12.14 | 12.28.0 |
| snowflake-connector-python | 3.12.4 | 4.3.0 |
| google-cloud-secret-manager | 2.19.0 | 2.26.0 |
| protobuf | 4.23.4 | >=6.33.5 |
| redis | 5.2.1 | 7.2.1 |
| neo4j | 5.28.1 | 6.1.0 |
| couchbase | 4.3.5 | 4.5.0 |
| faker | 33.1.0 | 40.8.0 |
| cryptography | ≥42 | 46.0.5 |
| oracledb | 1.4.2 | 3.4.2 |
| requests | ≥2.23 | ≥2.32.5 |
| pydantic_core | 2.27.2 | 2.41.5 |
| setuptools | 75.1.0 | 82.0.0 |
| typing_extensions | 4.12.2 | 4.15.0 |
| pymongo | ~4.3 | ~4.16.0 |
| collate-sqllineage | ~1.5 | 2.0.2 |
| google-cloud-storage | 1.43.0 | 3.9.0 |
| google-cloud-datacatalog | >=3.23.0 | 3.28.0 |
Required Developer Actions¶
-
Validate all HTTP integrations following the urllib3 2.x upgrade, with particular attention to custom SSL configurations.
-
Re-test PySpark pipelines to ensure compatibility with the target Spark version in use.
-
Verify that existing scikit-learn models load and function correctly with version 1.8.0.