Release Notes - What's new in Kyvos 2026.2.x
Here’s a list of Kyvos versions for Kyvos 2026.2.x releases:
Features and Enhancements for 2026.2.2
The Kyvos 2026.2.2 release brings the following new features and enhancements.
Support for Apache Impala Connectivity as a Data Source for No-Spark: With this release, Kyvos supports Apache Impala as a data source in No-Spark architecture, enabling users to leverage Impala data for processing semantic models.
Features and Enhancements for architecture2026.2.1
The Kyvos 2026.2.1 release brings the following new features and enhancements.
Support for Dedicated Compute Without Wildcard Access in AWS: With this release, Kyvos supports dedicated compute without wildcard access by removing the ‘autoscaling:DescribeAutoScalingGroups’ permission.
Note: Removing this permission results in specific functional restrictions. Refer to the Remove dependency of wildcard permission section for more details.Kyvos SCIM Provisioning with Microsoft Azure Entra ID using OAuth 2.0: With this release, Kyvos supports configuring System for Cross-domain Identity Management (SCIM) provisioning between Microsoft Azure Entra ID (formerly Azure Active Directory) and Kyvos using OAuth 2.0 authentication, enabling automated user provisioning and management.
Support for Sort Field on Parent Child Hierarchy (PCH) and Alternate Hierarchy in No-Spark: With this release, Kyvos now supports sort fields on Parent Child Hierarchy (PCH) and Alternate Hierarchy in MDX queries and Kyvos Visualization.
Features and Enhancements for 2026.2
The Kyvos 2026.2 release brings the following new features and enhancements.
Kyvos Dialogs Enhancements
This release includes the following enhancements to Kyvos Dialogs:
Support for Join Operations with Multiple Semantic Models: Kyvos Dialogs now supports querying across multiple Kyvos semantic models through AI Spaces. This enhancement enables AI-driven queries to span multiple semantic models, delivering unified insights and a more comprehensive conversational analytics experience.
Support for Verifying Question: This release supports to verify questions by marking (liking) them when the generated SQL accurately aligns with the user’s analytical intent. This enhancement provides a structured and governed mechanism to validate the accuracy of LLM-generated queries.
Enhancement in File Uploading: This release enhances the file upload workflow by detecting previously processed files and prompting users to select the associated AI Spaces. Users can choose to update existing data or initiate fresh analysis by creating new AI spaces, providing greater flexibility and improved data management.
Optimized Natural Language Query (NLQ) Token Consumption in Space: With this release, prompt optimizations have been implemented to reduce overall token consumption during Natural Language Query (NLQ) execution in Space. This improvement enhances performance compared to previous versions.
Note: The enhancement has been validated and certified with Azure OpenAI 4.1.
Kyvos Modern Architecture (No-Spark Ingestion) Enhancements
This release includes the following no-Spark enhancement:
Kyvos Lakehouse for AWS with Apache Iceberg Formats: Kyvos Lakehouse can directly read Apache Iceberg files stored in an S3 bucket and convert them into Kyvos-compliant Apache Iceberg files for semantic model processing. This support covers Amazon S3, removing the need for external catalogs or SQL engines.
Support for Azure Databricks SQL Warehouse as a SQL Engine: This release supports Azure Databricks SQL Warehouse as a SQL engine for Relational, Hybrid multidimensional and metadata-only semantic models.
Extended SQL JOIN Support: Kyvos now supports RIGHT OUTER JOIN and FULL OUTER JOIN queries in the SQL querying.
Snowflake Key Pair Authentication Support
Kyvos now supports an additional authentication mechanism as a key pair for Snowflake. This secure, text-based configuration provides long-lived authentication aligned with enterprise security standards and eliminates the need for frequent password rotation.
Custom Attribute Support with Azure Active Directory
This release supports custom attribute (mail) with Azure Active Directory (Azure AD) for Row-Level Security (RLS) after the semantic model is processed. This enhancement enables organizations to enforce access controls based on Azure AD custom attributes, providing greater flexibility and alignment with enterprise identity management policies.
PostgreSQL TLS Certificate Validation Enhancement
This release supports validation checks for PostgreSQL TLS certificates, including truststore.jks and keystore.jks files. PostgreSQL Certificate extraction and configuration are now performed only on the Kyvos Manager node in case of Install PG. For external repository, PostgreSQL certificate extraction and configuration are not performed on any node.
Enhancements to Kyvos Migration Utility
This release includes the following migration utility enhancements:
Added support to select a common Kyvos connection when creating a new semantic model through the utility.
Enhanced Power BI report handling for dynamic parameters created within workbooks.
Added support for custom SQL-based definitions while migrating Tableau workbooks.
Support for Special Characters in Measure Names
This release introduces support for special characters such as \, /, %, and () in base measure and calculated measure names during semantic model processing. This enhancement provides greater flexibility in naming conventions and improves compatibility with diverse reporting requirements.
SQL Expression-Based Support in Row-Level Security (RLS)
Kyvos now supports SQL expression-based Row-Level Security (RLS), enabling organizations to define custom SQL rule expressions to enforce granular data access controls. This capability allows access restrictions based on user-specific attributes such as division, job function, organizational unit, or signature category.
Upgraded Apache Tomcat Version
Kyvos and Kyvos Manager have been upgraded to Apache Tomcat 10.1.52.
Cloud Support Enhancements
This release comes with the following enhancements for AWS, Azure, and GCP clouds.
Time Zone-Based Scheduling: Kyvos now allows users to create and view schedules based on a configured time zone, eliminating the need for manual time conversion. When the configured time zone is modified, existing schedules are automatically adjusted to align with the newly selected time zone.
Support for Downloading Archived Heap dump, JVM Crash file, and Out file: With this release, users can view and download archived heap dump, JVM crash, and out files for the BI Server and Query Engine from Kyvos Manager under Monitor > Log Files > Archived.
Enhanced Filtering for Service Logs and Archived Diagnostic Files: This release improves Kyvos Manager (KM) by adding advanced filtering for listing and downloading service logs (by start and end date and time) and archived diagnostic files (by file type and date/time), including heap dumps, out files, JVM Crash, and log files.
Support for Single Node Standard (SNS) Automated Deployment: With this release onwards, Kyvos supports SNS automated deployment for all clouds.
AWS Enhancements
Support for Cluster Scaling Without EC2 Wildcard Access: With this release, Kyvos supports cluster scaling without EC2 wildcard access by removing the following IAM permissions:
ec2:DescribeInstances
ec2:DescribeInstanceStatus
ec2:DescribeNetworkInterfaces
ec2:DescribeInstanceTypes
ec2:DescribeInstanceTypeOfferings
Note: After removing the permissions, it has certain limitations. For more details about the limitations, see the Remove dependency of wildcard permission section.
Support for Private Deployment of Lambda Functions: With this release, automated deployments support private deployment of Lambda functions, which strengthens security, improves network control, and ensures compliance with enterprise infrastructure requirements
Support for AWS Kubernetes Version 1.34: This release supports EKS Kubernetes version 1.34 for new deployments and upgrades, as version 1.32 is approaching end of life (EOL).