25.3 enhancements and patches
Review the BMC HelixGPT 25.3 enhancements and patches for features that will benefit your organization and to understand changes that might impact your users:
| Version | SaaS | On-premises | Fixed issues | Updates and enhacements |
|---|---|---|---|---|
| 25.3.00 | ✅️ | Known and corrected issues | 25.3 enhacements | |
25.3 enhancements
Watch the following video (02:27) to learn about the latest enhancements in BMC HelixGPT 25.3:
HelixGPT Manager is rebranded to HelixGPT Agent Studio
HelixGPT Manager, the console that helps administrators to navigate and manage agentic AI capabilities across applications, is rebranded to HelixGPT Agent Studio. This new name for the administration console better reflects its purpose as the central workspace for configuring, managing, and optimizing Agents, Skills, Prompts, Tools, and Toolkits.
For more information, see User interface overview.
Enhanced incident resolution with BMC Helix Ops Swarmer integration for Microsoft Teams
The integration of BMC Helix Ops Swarmer with Microsoft Teams enables service desk agents to manage and resolve both BMC Helix ITSM and ServiceNow incidents within Microsoft Teams.
This enhancement bridges BMC Helix ITSM and BMC Helix AIOps to automatically correlate incidents with operational events, accelerating root cause identification and resolution.
- For information about BMC Helix Ops Swarmer agent, see BMC Helix Ops Swarmer.
- For information about collaboration for BMC Helix ITSM incidents, see Integrating BMC Helix ITSM and BMC Helix AIOps with Microsoft Teams for incident collaboration.
- For information about collaboration for ServiceNow incidents, see Integrating BMC Helix AIOps with Microsoft Teams for ServiceNow incident collaboration.
Create knowledge articles from incidents by using the Knowledge Curator AI agent for faster incident resolution
As a Service Desk Agent, create knowledge articles from incidents directly in BMC Helix Knowledge Management by ComAround by using the Knowledge Curator AI Agent.
This capability provides the following benefits:
- Users need not navigate away from the incident to create a knowledge article.
- Incident data, such as description, summary, and work notes is used to generate the knowledge article.
- The knowledge article is automatically pinned to the incident for ready reference.
Learn more about how to create knowledge articles by using the Knowledge Curator in BMC Helix ITSM in Creating knowledge articles from incidents by using Knowledge Curator.
Request a service through chat for a faster resolution
End users can request services through a conversational interface, experiencing quicker and more intuitive interactions. This enhancement streamlines the request process and helps users get faster resolutions without navigating complex wizards.
For catalog administrators, this reduces the effort needed to configure prompts manually, making it easier to manage large service catalogs while improving the overall user experience.
For more information, see Using BMC HelixGPT to make services requestable via chat.
Invoke an AI agent from the BMC Helix Innovation Studio workflow
In the Process designer, business analysts can add the Invoke Agent element to trigger an AI agent to perform a task. The Invoke Agent element provides the following benefits:
- Enhance automation by leveraging AI capabilities within workflows.
- Retrieve responses dynamically based on user input.
- Integrate conversational experiences seamlessly into business workflows.
For more information, see Invoking an AI agent from the process.
Improve response accuracy with predefined clarification options
Administrators can configure static clarification options by using the QueryClarification record in BMC Helix Innovation Studio. These predefined options appear as clickable buttons when user queries are unclear, helping users refine their input and receive more accurate responses.
For more information, see Configure the predefined clarification options.
Improve semantic indexing with a new embedding model
BMC HelixGPT supports the sentence-transformers/paraphrase-multilingual-mpnet-base-v2 model for semantic indexing. You can configure this embedding model for indexing to enhance document chunking and improve BMC HelixGPT’s ability to retrieve and reference relevant content. This model generates more precise and contextually accurate responses.
For more information, see Updating the embedding model for semantic indexing.
What else changed in this release
| Update | Product behavior in versions earlier than 25.3 | Product behavior in version 25.3 |
| Test prompts for image upload | Business analysts can test prompts that include image uploads. This capability improves prompt reliability and reduces production issues, especially for use cases that rely on visual input. | |
| Filter prompts based on applications | When you try to link a prompt to a skill, the Link Prompt dialog box displays a list of prompts for all applications.
| When you try to link a prompt to a skill, the Link prompt dialog box displays the list of prompts for the current application. For more information, see Linking a prompt to a skill.
|
| Share files in PDF or DOCX format while interacting with agents | AI Agents support document file types such as PDF and DOCX. With this capability, the Catalog administrators can upload documents during a conversation. For example, while updating the Catalog Profile section, administrators can provide files such as service descriptions, media assets, or SLA documents through the chat, which can then be uploaded to the respective attachment fields. For more information, see Managing AI agents. | |
| Seamless transfer to a live agent with the Live Agent Tool | Users can smoothly transition from a virtual assistant to a live agent through the Live Agent Tool, ensuring faster resolution for complex queries. For more information, see Managing AI agents. | |
| Access documents stored on Microsoft SharePoint | You can access documents stored on Microsoft SharePoint if you have permission to access them. This enhancement makes it easier to find the information you need without switching platforms. For more information, see Adding data sources in BMC HelixGPT. | |
| Test connections for AI service providers | The test connection is only supported for Azure OpenAI and OpenAI models with API type authentication.
| The test connection support is extended to Google and Oracle models with API type authentication.
|
| New LLM model support | BMC HelixGPT now supports the Amazon Bedrock Converse API. For more information, see Models in BMC HelixGPT. | |
| Feedback options in the Microsoft Teams channel | You can share feedback for the chat response in the Microsoft Teams channel. With this option, you can provide consistent feedback for improvement. For more information, see Enabling Microsoft Teams as a communication channel |
