Working with AI Space
This documentation is for Kyvos 2026.9.x. To view documentation for other Kyvos versions, click here. |
|---|
Note
This section applies to Kyvos Dialogs Quick Analysis and Deep Dive Analysis.
Overview
This section details how to use the AI Space in Kyvos to interact with data using natural-language queries and automatically generate insights, visualizations, and analyses.
AI Space in Kyvos enables conversational analytics across multiple data sources within a unified business context. It acts as a container that consolidates relevant entities for semantic models.
By organizing these entities within a single AI Space, Kyvos supports seamless conversational Q&A across multiple semantic models and business relationship graphs, enriched with external data sources. AI Spaces provide centralized AI settings, including shared business context, tool definitions, and business rules, ensuring accurate query interpretation, consistent SQL generation, and insights beyond simple “what” questions.
Creating an AI Space
On this page:
To create an AI Space, perform the following steps.
From the Toolbox, select AI Space.
Click the three dots (…) actions menu > Add Space.
Select the following from the Available Entities list.
Semantic Models: Select this option and the AI Space panel is shown, displaying a list of semantic models.
Business Relationships: Select this option and the AI Space panel is shown, displaying a list of business relationships.
In the Space Assets pane, drag and drop the semantic model and Business Relationship graphs to the Space Objects pane in the right.
Click Save. The new AI Space is ready.
Advanced settings
In the Advanced section, you can define the global LLM connections for the AI Space. Specify the LLM connection and Embedding connection for metadata, query, and summary generation.
Kyvos validates AI connections configured for an AI Space and its associated semantic model. If an AI connection is invalid, a warning is displayed on the AI Settings page and during semantic model validation, while still allowing you to save the configuration.
To configure global LLM connections for this space, perform the following steps.
Navigate to the Settings option.
In the Settings pane, enter the following details.
Area | Parameter | Description |
|---|---|---|
General | Allow Sending Data to LLM | Configure whether data from the AI space is sent to the LLM by selecting one of these options:
|
Query Generation | LLM Connection | Select the LLM connection used by Kyvos Dialogs and Kyvos Viz. for interactions with this semantic model. |
| Embedding Connection | Select the embedding connection to generate embeddings of AI Space metadata for semantic search and retrieval. Helps the system understand context, match similar queries, and improve result accuracy. |
Summary Generation | LLM Connection | Select the LLM connection used for natural language generation on this semantic model. |
Max Data Points for Summary | This property configures the maximum number of data points involved in generating the summary. | |
Metadata Generation | LLM Connection | Select the LLM connection from the list. This connection automatically generates metadata such as descriptions, tags, and business context for data models. |
More Actions
Click the Actions menu (…) and do the following:
Validate the space
Share the space via link or grant access.
Refresh the space to get the updates.
Add Note or Tag for the space.
Click Save. The Space will be added and displayed to the AI Space panel.
Sharing Semantic Models
Semantic models of an AI space can be shared with users or groups by granting the appropriate access permissions.
Note
From Kyvos 2026.9 onwards, the following enhancements ensure that users can run queries in Kyvos Dialogs only on the semantic models for which they have the required permissions:
AI Space Access Validation: Enhanced access validation for AI Spaces to ensure that users can query an AI Space only when they have access to at least one of its associated semantic models. A clear access-related message is now displayed when the user does not have access to any semantic model in the selected AI Space.
Restricted Semantic Models in Feedback: Improved access control in the feedback flow by excluding semantic models that the user is not authorized to access. This prevents restricted semantic models from appearing as selectable entities or being included in feedback-related queries.
Restricted Columns and Values: Columns and data values belonging to semantic models that the user cannot access are no longer displayed, ensuring that restricted information is not exposed through these supporting features.
Semantic Model Listing Across Integrations: Enhanced privilege-based filtering of semantic models listed through MCP, Microsoft Teams, and Slack. Users can now see and select only the semantic models they are authorized to access when interacting with an AI Space through these integrations.
To share a semantic model in an AI Space with a user or group, perform the following steps.
Navigate to the Toolbox > AI Space and select the appropriate AI space.
Under Semantic Models, select the appropriate semantic model from the list and then click the AI Settings link for that model. The semantic model page opens.
Click the Actions menu (…) and select Share.
Select one of the options:
via Link: Generates a shareable link that can be used to grant the specified permissions to a user or group.
Grant Access: Opens a dialog box where you can select users or groups and assign access permissions for the semantic model.
Select the user or group with whom you want to share the semantic model. Under the Share With section, drag and drop the user or group into one of the following permission categories: Read Only, Read and Write, or Deny.
Click Share. The semantic model is shared with the selected user or group with the assigned permission(s).
Collaborative editing support
From Kyvos 2026.5 onwards, Kyvos supports collaborative editing by allowing multiple users to work on the same AI space model simultaneously without enforcing edit locks. It provides visibility into active users, detects concurrent modifications, and offers conflict resolution and merge capabilities to help users manage and reconcile changes efficiently.