Aisera Product Release Notes - June 16, 2026
Review what's new and changed, and the fixes and limitations for the Aisera platform in the June 16, 2026 release.
Conversational AI
What's New
Contextual Information for Fallback Answers in ConvAI 2.0 Applications/Bots
The ConvAI 2.0 version of Aisera application/bots now provide intelligent, context-aware fallback responses when the requested information cannot be retrieved.
Instead of returning a generic fallback message, the application/bot analyzes the user's intent and responds with a more helpful explanation based on the situation:
No answer available
Access-restricted information
Configurable response format
This enhancement improves transparency, reduces user confusion, and ensures users receive the most relevant guidance available, even when a direct answer cannot be provided.
To enable contextual information for ConvAI 2.0 fallback messages:
Navigate to Settings > AiseraGPT in the Aisera Admin UI.
Verify that your application/bot is set to use Conversational AI v2.0.
Note that intents for Conversational AI 2.0 applications are detected using the Title or Description field in a Workflow or Hyperflow.
Select the application/bot that includes the messaging that you want to modify.
Choose the Advanced tab on the application/bot Edit AiseraGPT window.
Select the checkbox next to the Add contextual information fallback answers option.
Flows (Workflows and Hyperflows)
What's New
Consolidated Hyperflow configuration management
Hyperflow tenant-level settings now live in one place: Settings > Configuration > Hyperflow. Previously, these settings were spread across multiple tabs, which made administration harder.
Tenant Level Hyperflow Configurations
Tenant-level model configuration for Hyperflows
You can now set default LLMs for Hyperflow execution and testing at the tenant level. This centralizes model selection across the organization while still allowing app-specific overrides when needed.
App-level model configuration for Hyperflows
You can now set the LLM used for Hyperflow execution and testing at the application level. This give you flexibility to optimize performance by selecting the most appropriate model for each bot or application.
This configuration is located at the top of the AI Automation > Hyperflows page. The configuration will only be visible if you have disabled the Enable Hyperflow V1 configuration. This configuration can be found under Settings > Configurations > Hyperflows
Bulk flow registry operations
You can now activate, deactivate, or delete multiple Workflows or Hyperflows simultaneously from an application. When activating or deactivating, the available bulk action updates based on the current selection all active flows show a deactivate option, all inactive flows show an activate option. Mixed-status selections do not support bulk activation or deactivation.
What's Changed
Enhanced workflow function export/import alignment
To ensure consistency with standard Workflow export/import behavior, Workflows used as functions (tools) in Hyperflows now export their last published version.
When importing to a target Application or Tenant, if a Workflow with the same name already exists, the exported version is imported as a Draft on top of the existing version history, preserving all previous versions while safely adding the new functionality.
Enhanced validation for application-level workflow publishing
The system now validates that required metadata, description and category, is complete before allowing imported workflows to be published in 2.0 applications.
Description enables conversational triggering and category is required for organization. Workflows missing either field are block from publishing.
Validation applies only to application-level workflows imported from 1.0 applications to 2.0 applications.
HTTP node variable selection
HTTP node functionality has been aligned with other nodes. Headers and Body configuration now supports easy selection from existing variables, resulting in a faster and more intuitive node building process.
OAuth variable handling in flow execution
OAuth user input variables are now securely processed as sensitive data, remaining encrypted during execution and masked in execution logs, channel messages, and debug information.
Fixes
Corrected Workflow import behavior to ensure repeated imports of the same Workflow overwrite existing Draft versions rather than reverting Published versions to Draft status.
Fixed an issue where a performance regression introduced in the previous release caused significantly slower workflow import times.
Ingestion and Connectors
What's New
Desk Connector Available in the Aisera Admin UI
Introducing a new out-of-the-box FreshDesk Connector that replaces the previous generic/override-based workaround. This productized connector natively integrates with FreshDesk and supports the full Knowledge Base Article hierarchy, including sub-folders.
This pre-mapped connector provides a scalable, maintainable, and fully supported integration path for FreshDesk customers with multi-level knowledge base hierarchies (such as Category > Folder > Subfolder > Article).
This new connector includes the following capabilities:
KB Learning - Full discovery path: Categories > Folders > Subfolders > Solution Articles. Supports filtering by status and visibility fields.
User & User Profile Learning: Ingests FreshDesk Agents data.
Ticket Learning: Ingests FreshDesk Tickets with date-range filtering support via DS Configuration.
For detailed information about how to use the FreshDesk connector, see FreshDesk Connector.
Disable Archiving Options for Incremental Data Ingestion
Incremental fetches do not see "old but still valid" docs, so these remain in a 'to be deleted' state and get archived, even though they still exist in the external system.
For example, in testing, this resulted in a drop from ~1k+ docs down to 200+ docs in SOR during incremental runs. Switching to a full (on-demand) fetch restored all docs.
Now you can set an option on your Data Source Configuration to prevent the incremental data ingestion from archiving 'old but still valid' article.
To set Disable Archivng Option:
Navigate to Settings > Data Sources.
Choose the Data Source that you are ingesting data from on an incremental schedule.
Select the pencil icon to open the Edit Data Source window.
Enable the checkbox for Disable Archiving.
This will ensure that incremental sync jobs do not archive or delete Knowledge Article documents simply because they were not updated in the source.
Incremental Data Source Schedule
What's Changed
The Box Connector now supports extracting metadata (metatags) from .boxnote files during ingestion
This enables the RAG layer to filter and serve knowledge articles based on attributes such as region, audience, access level, content type, and other custom metadata fields configured in the Box Data Source.
Key Capabilities of Metadata Extraction: Metadata Extraction: A new "Fetch Metadata" checkbox is available in the Box integration configuration. When enabled, the connector calls Box's metadata API endpoints to retrieve all metadata templates and their field values attached to .boxnote files.
Downstream Propagation: Extracted metadata is included in the ingestion payload under a customMetaData object, preserving the full template structure (e.g., template name, field keys, and values).
ACL-Based Filtering via Custom Scripts: Administrators can apply custom transformation scripts in the field mapping configuration to map specific metadata fields (e.g., region) to tagKeys and tagValues. This enables attribute-based access control and filtered retrieval at query time.
Multi-Value Support: Array-type metadata fields (e.g., a file tagged to multiple regions like ["AMER", "EMEA"]) are correctly parsed and mapped as individual tag entries.
To Enable Metadata Fetching:
Navigate to Settings > Connectors in the Aisera Admin UI.
Open the Box Connector Data Source.
Select the pencil icon to open the Edit Data Source window.
Choose the Configuration tab.
Enable the Fetch Metadata checkbox.
Configure field mappings to map your desired metadata fields to tagKeys/tagValues using a custom transformation script.
Enable auto-commit on the datasource and attach it to the target bot.
Validate that metadata appears correctly in the System of Record (SOR) and that retrieval respects the configured attributes.
Issues Resolved in the Release
The following table includes bugs that were fixed in this release.
Aisera ID
Product Area
Summary
6.16.2026.1
Agent Assist
Agent Assist w/Zendesk widget is not scrollable to the bottom when wrap up card is available
6.16.2026.2
Agent Assist w/Zendesk prediction cards are appearing as collapsed irrespective of the widget config settings
6.16.2026.3
In Agent Assist w/Salesforce, the Next button should be disabled or greyed out upon reaching the final page
6.16.2026.4
Discrepancy in the value of Analyzed and Agent action values in the 'Agent Assist Aggregate over time' graph
6.16.2026.5
Webchat, Chatbot, Copilot, and Universal Bot
Screen repeats "Message from bot" multiple times before displaying the actual message
6.16.2026.6
Webchat: Privacy Policy Message is not translating for Webchat Channel
6.16.2026.7
Chatbot: while using slack sdk bolt, links are rendered incorrectly
6.16.2026.8
Chatbot: Missed link rendering when img appear inline and error on switching between normal and bolt mode during same session
6.16.2026.9
Chatbot: Slack even after updating the tenant level flag the switch to new sdk is not happening
6.16.2026.10
AI lens is failing to load - If co-pilot channel and webchat channel co-exist
6.16.2026.11
Chatbot: Conversation Action Labels Using App Language Instead of Chat Language
6.16.2026.12
Copilot: Citations Fix active badge color, sidebar rendering, and post refresh pill rendering
6.16.2026.13
Names not shown in UI for copilot channel
6.16.2026.14
Campaigns
Form Submit is failing in conversation server in Campaigns
6.16.2026.15
Campaign flows fail validation with "missing a description" error when publishing under a Conversational AI v2.0 bot
6.16.2026.16
Connectors
Known credential Expiration Date Field should not be a mandatory field In Create/Edit Integration Page
6.16.2026.17
Datasource Execution Summary page is not showing graph section
6.16.2026.18
Error on field change - FileData
6.16.2026.19
Flows (Workflows and Hyperflows)
Huge delay with Flow import process after 5/15 release
6.16.2026.20
Repeated Flow Import Incorrectly Converts Published Version to Draft Instead of Overriding Existing Draft [No Changes Are Made Before Publishing]
6.16.2026.21
Notification
Credential Expiration date is not saved as expected in DB while creating Notification setting for a specific Integration
6.16.2026.22
Credential Notification Emails: Fix formatting of special Characters
6.16.2026.23
Credential Notification Screens Issues
6.16.2026.24
User Profiles and Roles
Credential Notification Emails: Fix formatting of special Characters
6.16.2026.25
Credential Notification Screens Issues
6.16.2026.26
User Profile Service: Prevent Duplicates Values in Group Names List
Last updated
Was this helpful?