Release Notes - What's new in Kyvos 2026.1.x

Here’s a list of Kyvos versions for Kyvos 2026.1.x releases:

Kyvos 2026.1 - Features and enhancements

Features and Enhancements for 2026.1

The Kyvos 2026.1 release brings the following new features and enhancements.

Kyvos Dialogs Enhancements

This release includes the following enhancements to Kyvos Dialogs:

Summary Streaming Support: This release introduces an LLM connection property (Enable Summary Streaming) to control summary streaming in Kyvos Dialogs, enabling summaries to be displayed progressively rather than all at once.
Support for Verified Sample Questions: This release supports adding verified sample business questions in natural language for AI Spaces and Semantic Models. These predefined questions help improve the LLMs and conversational analytics experience with Kyvos Dialogs. Additionally, you can also import or export sample queries.
Support for Generating Follow-Up Questions: This release adds support for generating relevant follow-up questions, allowing users to explore data and identify root causes.
Note: You must enable key insight generation to use this feature.
Automatic Metadata Generation Support Using Kyvos Copilot: This release supports using Kyvos Copilot to automatically generate business context and tags for semantic model columns, improving metadata enrichment and usability.
Social Media Login Support in Web-Based Kyvos Dialogs: This release supports log in to web-based Kyvos Dialogs using social media accounts through Kyvos Public Server.
UI Enhancements for AI Settings: This release improves the AI Settings UI as part of ongoing updates, including:
- Added separate tabs to manage general settings (LLM configurations), column metadata, and context.
- Improved tag and business context readability.
- Refined key element management with more precise show/hide controls for a more intuitive experience.

Kyvos Modern Architecture (No-Spark Ingestion) Enhancements

This release includes the following no-Spark enhancement:

Kyvos Lakehouse for AWS with Parquet Files: Kyvos Lakehouse can directly read Parquet files stored in data lake folders and convert them into Kyvos-compliant Parquet files for semantic model processing. This support covers Amazon S3, removing the need for external catalogs or SQL engines. Activate this feature to read Parquet files and process the semantic model.
Writeback Enhancements: This release introduces enhancements to Writeback, including support for dimension-based filtering on semantic models. Additionally, Writeback now supports unrelated dimensions in Kyvos views or filters.
Note: Measure-based filtering is not supported for Writeback.
Support of Incremental Identifier: This release supports incremental identifier in the no-Spark ingestion.
Support for Relational Multidimensional Model Querying on BigQuery with Metadata-Only Processing: This release supports Relational Multidimensional model querying on BigQuery using Metadata-Only semantic model processing.
Support for OR Operator in Row-Level Security (RLS): This release introduces support for the OR operator in Row-Level Security (RLS) rules for related dimensions and the fact table in a no-Spark ingestion. The functionality is controlled by the RLS_OR_SUPPORT (biserver.properties) property, which is enabled by default and set to true.

Support of Natural Logarithm and Standard Deviation MDX Functions

With this release, Kyvos has added support for LN (natural logarithm) and STDEVP (Standard Deviation) MDX calculated measure functions.

LN Function: Enables users to compute natural log values within custom expressions and measures, supporting scalar, mathematical, and semantic model-based expressions with standard logarithmic behavior.
STDEVP Function: Enables users understand data variability and consistency across time periods, dimensions, and business segments.

Enhancements to Kyvos Migration Utility

This release includes the following enhancements to Kyvos Migration Utility:

Added UI support to allow users to provide folder names and Kyvos connection names on the semantic model creation screen.
Introduced UI support to review generated summaries as part of the summaries feature.
Enhanced Power BI report support with updated UI text and behavior specific to Tableau and Power BI.
Implemented UI improvements, including copy-to-clipboard functionality and row-level actions.
Added support for joins and single-table semantic models with SQL-based definitions.

Cloud Support Enhancements

This release comes with the following enhancements for AWS, Azure, and GCP clouds.

Support for PostgreSQL 16 Support in Marketplace and SaaS: This release supports PostgreSQL 16 (installed as a service) for Marketplace (Azure Marketplace Free and Paid) and SaaS (AWS SaaS SNS) deployments.

AWS Enhancements

Support for RDS Repository Operations Without Wildcard Access: With this release, for the Kyvos RDS repository, users can take a pg dump based back up of the repo state during upgrade and while rollback, the repository state is restored using the PostgreSQL dump captured during the upgrade or RDS snapshot, depending on how the database backup was initially created for the target rollback state.
Note: Only RDS snapshot-based backup is discontinued. Rollback by creating RDS from a snapshot is still supported and requires permission.
Support for Deleting Secret Keys and Values from Secret Manager on Connection Deletion: With this release, deleting a connection from Kyvos also removes the associated secret keys and values from Secret Manager. To enable this behavior, the IAM user must have the secretsmanager:PutSecretValue permission.
Support for Using Pre-Created Kubernetes Clusters in Automated Disaster Recovery: Users can now use a pre-created Kubernetes cluster during automated disaster recovery to reduce recovery time. In the dr_Single_Template.json script, select Kubernetes_Precreated during resource recovery to enable this support.
Managing Kubernetes Configuration Preferences from Kyvos Manager: This release supports modifying Kubernetes configuration preferences directly in Kyvos Manager, eliminating the need for terminal access or manual downloading and uploading of Kubernetes configuration preference JSON files.

Azure Enhancements

Multiple Compute Server Cluster Support: Kyvos now supports multiple compute server clusters, each associated with a dedicated Kubernetes (K8S) cluster or user pool. Users can define multiple compute server cluster connections and select the appropriate connection during semantic model processing configuration. This provides flexible resource management, ensures workload isolation, and allows the high-priority semantic model to run independently.
Filtering Kyvos Manager Operations by Subtype: Kyvos Manager now supports searching and filtering operations by subtype. This subtype applies only to two Kyvos Manager operations: Cloud Action and Node Action. This support simplifies the analysis of cluster operations through search and filter.
Extended Support for RHEL 9 in Azure SaaS and Marketplace deployments: Kyvos supports Red Hat Enterprise Linux (RHEL) 9 for new SaaS and Marketplace deployments.
Exporting Data for Databricks SQL Warehouse to Storage Container in User’s Specific Account: Users can export data during semantic model processing to a storage container in their specified directory instead of Kyvos-owned storage.
Note:
- To enable this, you must specify the required properties during the Azure Databricks SQL connection.
- If no export directory is specified, export defaults to a Kyvos-owned directory in Kyvos’s account/tenant.

Kyvos Manager Enhancements

Support for Logging Process ID (PID) of Launched Process (BI Server and Query Engines) in Operation Logs and Audit: Kyvos Manager now displays the Process ID (PID) of BI Server and Query Engines services in all the operation logs where these services are started. The PID for these services is also displayed in the audit logs, aiding in troubleshooting and analysis.
Support for Downloading BI Server and Query Engine Local Diagnostic Files (Heap Dump, JVM Crash, and Out Files): With this release, users can view, delete, and download heap dump, JVM crash, and out files for the BI Server and Query Engine directly from Kyvos Manager under Monitor > Log Files.
Support for Archiving Heap dump, JVM Crash file, and Out file to Distributed File System (DFS): With this release, after starting the BI Server and Query Engine services, the heap dump, JVM crash file, and out file will be purged from the local and archived to DFS under the logs folder.