4.99¶
Update Okta SSO Settings
Due to security updates, you need to change the Okta SSO settings. On the Okta side, go to the SAML settings and replace the value in the Audience URI with the URL configured in idpIssuer
. For details, refer to the Okta guide in the Cognigy Help Center.
Upcoming Feature Deprecations and Removals
On April 29, 2025, we will discard say
verbs with empty text content
affecting the following Voice Gateway Nodes: Say, Question, and Optional Question.
Refer to the list of all upcoming removals on the Deprecations and Removals page.
4.99.0¶
Cognigy.AI¶
Released April 15, 2025
Features¶
Free of Charge Model for Building AI Agents¶
Introduced a platform-provided LLM that allows users to build and test AI Agents without additional licensing costs. Applicable only to design-time features.
MCP Tool for AI Agents¶
Introduced AI Agent child Nodes: AI Agent MCP Tool and Call MCP Tool. Along with tool actions, you can use Model Context Protocol (MCP), which boosts tool actions by providing a standardized connection between AI Agents and external services. It allows smooth integration with remote systems like APIs and databases, enabling the AI Agent to perform more complex tasks in a secure and efficient way.
Mistral AI – New LLM Provider¶
Added support for Mistral AI as a Large Language Model provider, expanding available options for LLM-driven AI Agents.
New Google Models Available¶
Introduced support for Google's latest models: gemini-2.0-flash
and gemini-2.0-flash-lite
, enabling faster, more efficient AI Agent interactions.
Improvements¶
- Renamed the Queue field to Queue ID in the Genesys Open Messaging handover provider settings
- Renamed the Create & Configure LLM button to Create in the AI Agent wizard
- Improved the warning system about migration in the Endpoint editor and in the new handover providers interface
- Changed the maximum file size for uploading Intent Trainer records from 100 MB to 150 MB
Bug Fixes¶
- Fixed the issue where service names were misaligned in the handover provider list of the Handover to Agent Node
- Fixed the issue where the last modified date of the Project and its resources wasn't updated when importing a Package or restoring Snapshots
- Fixed the issue where missing labels weren't added for Knowledge Stores for the new models
- Fixed the issue where the Locale ID wasn't validated before switching locales during Flow execution
- Fixed the issue where the handover service by ID returned a
404
error even when a valid ID was provided - Fixed the issue where timeout options were ignored for AWS Bedrock, Google Gemini, and Vertex AI models
- Fixed the issue where exporting a large number of entries in the Intent Trainer used excessive memory, leading to out-of-memory (OOM) errors
- Fixed the issue where closing Say and Question Nodes in the UI triggered an incorrect
unsaved changes
warning - Fixed the issue in the Flow editor where Extensions couldn't be selected from the All tab in the Extensions menu
- Fixed the issue where the RingCentral Engage handover didn't occur when created with the new handover settings
- Fixed the issue where double hyphens in the domain name weren't allowed by the URL validation
Cognigy Voice Gateway¶
Released April 15, 2025
Improvements¶
- Added the Enable Immediate Hangup parameter to the Hang Up Node. Activating this parameter immediately ends the call when a Hang Up Node is triggered, bypassing active operations. This parameter is useful for automatically ending calls when an answering machine is detected
- Improved
api-server
by adding support for thenova-3
model provided by Deepgram. You can now specify this model in the following Nodes:- Set Session Config Node. Go to the Recognizer (STT) settings, select Deepgram as the vendor, then select Custom from the Deepgram Model list
- Nodes with Set Activity Parameters and Voice Gateway channel, such as Say, Question, and Optional Question
- AI Agent Node. Go to the Voice settings, select Deepgram as the vendor, then select Custom from the Deepgram Model list. Alternatively, you can specify the model in the AI Agent wizard
Bug Fixes¶
- Fixed the issue where the
CALL_COMPLETED
event wasn't sent because the WebSocket connection was closed beforehand - Fixed the issue in the Voice Gateway Self-Service Portal where the synthesizer language for Deepgram and ElevenLabs wasn't updated correctly. As a result, the UI showed the default language instead of the selected one
Cognigy Webchat¶
Released April 15, 2025
Improvements¶
- Added parameters for activating the deletion of the current conversation or all conversations to the Webchat v3 Endpoint settings
- Added support for progressive rendering of mixed text, AI streams, and rich messages in order, with updated quick reply behavior and full backwards compatibility
- Updated Webchat v3 to v3.18.0
Bug Fixes¶
- Fixed the scroll button's behavior in the input area to prevent overlap in dynamic layouts
Cognigy Live Agent¶
Released April 15, 2025
Improvements¶
- Added a warning about cross-origin resource sharing (CORS) compliance to Account Settings > Notification Sound > Custom
Bug Fixes¶
- Fixed the issue where the conversation filters didn't show the Webchat v3 option
- Fixed the issue where deleting large inboxes caused request timeouts
Cognigy Insights¶
Released April 15, 2025
Improvements¶
- Changed the pre-aggregation error notification so that it doesn't block the loading of other charts
Bug Fixes¶
- Fixed the issue where the Locale and rating filters were applying to only one value. Also fixed the rating filter for the Unique Contacts chart
Infrastructure¶
Cognigy.AI + Insights + AI Copilot¶
cognigy-auth-jwt¶
A new Kubernetes secret, named cognigy-auth-jwt
, has been introduced in this release of Cognigy.AI.
This secret will store a 128-byte random hex value, which will be used for secure JWT (JSON Web Token) authentication.
The key is essential for ensuring the integrity and confidentiality of authentication processes within Cognigy.AI, providing a secure mechanism for handling authentication tokens in the Cognigy.AI platform.
Cognigy Voice Gateway¶
Updated the credentials handling of Homer:
- Added property
{{ .Values.homer.enabled }}
. The default value istrue
(recommended) - If the property is set to
true
, a new secret,vg-homer-password
, is created during the deployment of the Helm chart - A new job,
homer-password-update
, has been added. This job updates the default admin user in Homer and updates the password based on the created secret
Version Compatibility Matrix¶
The version compatibility matrix was updated for the following Cognigy products: