Introduced Troubleshooting Agent

New Feature

The Troubleshooting Agent helps identify how similar problems were resolved in the past by surfacing relevant examples and required parts from prior work. It analyzes historical service data to present related cases, showing what failed and how the issue was addressed. This enables technicians to compare past outcomes and make informed decisions, while the agent itself supports comparison only and does not automate decisions.

Diagnosing complex issues often requires technicians to search across multiple sources, including service history, documentation, and other reference materials. This slows resolution and can lead to repeat visits. The Troubleshooting Agent shortens that cycle by making prior experience visible during active work. Teams spend less time searching solutions, and fixes rely on evidence from past outcomes.

Administrators configure which work orders the Troubleshooting Agent analyzes and how relevance is determined. Configuration defines the problem details, solution details, and asset information used for matching similar cases. Administrators also control filters that limit comparisons to appropriate assets or time ranges. Once enabled, the agent builds and maintains its reference data from completed work orders.

Administrators can define a new AI action for the Work Order object by referencing the Troubleshooting Agent in the AI Instructions and providing relevant guidance. Administrators can use the Run AI Action to troubleshoot open Work Orders directly from the ServiceMax AI Console. For more information, see Enabling Troubleshooting Agent and Troubleshooting Agent Configuration.

AI Chat support in Windows Go

New Feature

AI Chat is available in Windows Go, providing in‑app access to ServiceMax AI. AI Chat launches directly from supported records, such as Work Order, Installed Product, Case, Location, and Account, and uses the active record as context for generating responses. Responses reference record data, link to related ServiceMax records, and surface relevant knowledge within the Windows Go app.

AI Chat in Windows Go provides in‑context access to ServiceMax AI across supported records. This capability keeps guidance inside the workflow and helps reduce manual searching and context switching.

Administrators control access to AI Chat in Windows Go through existing AI enablement settings. Enabling AI Chat for a Windows Go profile exposes the chat entry point in the application. Access depends on permissions to at least one AI agent. No Windows‑specific configuration is required, and management continues through the AI Console.

End users can interact with AI Chat directly in Windows Go and start conversations from supported records, such as Work Order, Installed Product, Case, Location, and Account. The active record is passed as context to the AI assistant. Responses include links to related records and source documents for quick reference. For more information, see Using ServiceMax AI Chat in Windows Go.

Capture Knowledge articles from Zinc Hotline chats

New Feature

Zinc AI Knowledge Capture extracts useful information from Zinc hotline chats and converts it into Knowledge Memos. It generates draft knowledge articles that users can review, edit, and save through the Knowledge Access agent. Saved entries are then available to the Knowledge Access agent to help answer similar questions in the future.

Expert knowledge is often shared in one-off conversations and is not captured for reuse. Zinc AI Knowledge Capture converts hotline chats into Knowledge Memos. These are accessed through AI Chat, where the Knowledge Access agent references relevant memos to help answer user questions, reducing repeat queries and dependence on individual experts.

Administrators control where and how knowledge capture runs. Configuration in the Zinc administration console allows administrators to enable or disable AI knowledge capture per hotline. For more information, see Managing Zinc Hotline Summaries.

End users such as Subject Matter Experts working in Zinc hotline requests can generate knowledge directly from an active or closed conversation. A Generate Knowledge Article option creates a draft from the chat content. The saved knowledge becomes available to ServiceMax AI, allowing future questions to surface the same guidance without reopening the original conversation.

AI run trace log from AI Usage

Enhancement

AI Usage now supports viewing detailed trace logs for individual AI runs. This enhancement extends visibility beyond summary‑level information by exposing execution‑level details, enabling deeper insights on run behavior and execution flow.

Administrators can view the trace log information for the specific AI run in the ServiceMax AI Analytics dashboard. They can also manually copy the log based on their requirements. The log contains technical trace information for that specific run and supports deeper analysis during troubleshooting or when sharing details with ServiceMax support. For more information, see Viewing Run Trace Logs.

AI Console navigation and usability improvements

Enhancement

The AI Console includes navigation and usability enhancements that improve readability, interaction, and interpretation of run details. These changes simplify movement between console pages and enhance clarity when reviewing AI activity.

Administrators can open AI Console pages in a new browser tab using standard right‑click actions, making it easier to switch between views. AI Action responses in the usage details view display clearer formatting, with readable line breaks and clickable record names. Visual indicators also improve clarity when reviewing allowed transactions. These changes reduce navigation effort and improve readability without requiring any new setup.

AI agent skills composer for Service Insights agent

New Feature

The AI agent skills composer is a simple interface to create, edit, and manage skills for the Service Insights agent. This gives a flexibility to add new skills in addition to the available out-of-the-box (OOTB) skills. These skills define targeted tasks, such as calculating metrics or querying data.

The skills composer ensures accurate, consistent AI behavior by letting ServiceMax AI leverage reusable, deterministic code. This feature enables scalable AI configuration while maintaining consistency, reliability, and administrative oversight.

Administrators can create, edit, and manage skills for the Service Insights agent. Instructions are authored in plain language and translated into a generated plan that uses deterministic logic, such as database queries and executable code. Administrators must review the generated plan before saving skill. The workflow supports iteration. Administrators can refine instructions, regenerate plans, and review results. Existing skills can be edited or deleted, except when referenced by an AI Action. Administrators also edit the OOTB skills using the same process. For more information, see Configuring AI Agent Skills.

End users receive consistent and reliable responses from the Service Insights agent when performing common analytical tasks. The agent applies the same logic each time it runs a skill. Results do not vary between executions and calculations remain accurate. For more information, see Using Agent Skills.

AI Action in‑page component for record pages

New Feature

AI Action in‑page component displays AI Action results directly on a record page. The component runs a selected AI Action and renders the response within the page layout on supported Salesforce records. Information appears in context, without requiring navigation to a wizard or separate screen. The component can run automatically on page load or on demand.

The AI Action page component ensures users get timely, in‑context information directly within the record view, with results shown in clear text and structured layouts. Service organizations need faster access to insights without interrupting established workflows. Embedding AI results directly in record pages improves visibility and adoption while reducing unnecessary navigation and clicks.

Administrators can add the AI Action in‑page component to supported record pages and select the AI Action to run. They can configure and control whether the action runs automatically on page load or requires a manual trigger. For more information, see Adding the AI Action Page Component and Adding the AI Action Page Component for Community Users.

End users see AI Action output immediately after opening a record. The component displays a loading indicator while the action runs, then shows the response inline. When users open a record, the component immediately displays AI‑generated results, without needing Service Flow Manager (SFM) wizards or manual steps.

AI Actions support in Windows Go

New Feature

AI Actions are now available on Windows Go. Service technicians can launch AI Actions directly from Service Flow Manager (SFM) wizard steps and view the response within the Windows Go application. The experience mirrors core AI Action behavior already available on other supported platforms.

Windows Go technicians previously lacked access to AI Actions during guided workflows. This gap limited access to contextual insights and forced reliance on manual steps or other devices. AI Actions support in Windows Go closes that gap. Technicians can now trigger AI Actions from SFM wizard steps and receive AI‑generated responses while working in the same workflow.

Administrators can configure AI Action–type SFM wizard steps for Windows Go using existing SFM activation rules. The same criteria that control AI Action visibility on other platforms apply in Windows Go. No additional configuration is required to enable record links or Source document references in AI responses. The system displays these automatically when present in the output.

End users can launch AI Actions from supported SFM wizard steps in Windows Go. The application displays a loading indicator while the action runs, then presents the response inline. Record names in the response appear as links for quick navigation. Knowledge Access results include links to source documents, allowing technicians to review supporting material without leaving the workflow. For more information, see Using AI Actions in Windows Go.

Account‑based AI actions and responses

New Feature

Account‑based AI actions and responses extend ServiceMax AI to support Account as a base object. AI Actions can now analyze and return insights across all locations, assets, problems, and service activity linked to an Account. The feature provides a consolidated, account‑level view of service data without changing existing workflows.

Service teams often review service history one record at a time. That approach fragments the customer story and hides patterns across locations and assets. Account‑based AI actions and responses allow AI agents to evaluate data at the Account level. The system correlates related records and returns a unified view of service experience and asset operations.

Administrators can configure AI Actions that use Account as the target object. The Service Insights agent supports the Account as an object type for AI Actions and includes it in its default data sources. Existing guidelines for Service Insights support Account‑scoped queries. For more information, see ServiceMax Data Sources Configuration and Configuring AI Actions.

End users can run AI Actions that answer Account‑level questions. Responses include insights across related locations, assets, and service visits. For more information, see Using Run AI Actions.

Document applicability rules for Knowledge Access agent

Enhancement

Knowledge Access agent now uses document applicability rules to control which documents apply to each product. The agent searches only documents that match the product context. This change improves accuracy in AI responses.

Administrators can link documents to a Product, Product Family, or Product Line. They can add or remove tags directly from the documents table. Additionally, they can update tags during document selection and search for products without leaving the page. Administrators manage document applicability efficiently through inline and bulk actions. The full list of applicability rules remains visible for each document.

End users see correct source documents in the AI response. The enhancement supports bulk-add and bulk-remove actions for applicability tags. This reduces the time required to configure the Knowledge Access agent.

For more information, see Adding Applicability Tags to Documents and Removing Applicability Tags from Documents.

SFM AI transactions support Header and Child Line creation

Enhancement

ServiceMax AI now supports Service Flow Manager transactions that create child line records. The AI can save a header record and its related child records in one action. This support expands where AI actions apply and reduces manual data entry.

Service technicians can complete tasks that require both header and child records in one request. This reduces rework and follow-up steps. Administrators can reference these transactions while configuring AI actions. No workflow change is required for end users. For more information, see Service Flow Manager Agent Configuration and Configuring AI Actions.

Enhanced AI error responses for clear guidance

Enhancement

ServiceMax AI now returns clearer error messages when it cannot complete a request. Each message explains the reason for the failure and points to a practical next step. This change helps end users understand missing information and modify their questions to retrieve desired AI response.

End users see specific explanations instead of generic failures. When required data is missing or incomplete, ServiceMax AI explains where it searched and what it could not find. Service Insights now clarifies empty fields, missing related records, or inaccessible fields such as List Price. Knowledge Access distinguishes between products with no documents and cases where no relevant results were found. End users receive clearer guidance on rephrasing questions, seeking administrator help, or using alternate sources, reducing confusion and increasing trust in AI‑powered workflows. For more information, see AI Insight Data Unavailability.

Introduced MCP server

New Feature

The Model Context Protocol (MCP) Server enables external AI agents to access ServiceMax AI capabilities. Administrators can use the MCP Server to whitelist AI Actions in the AI Console, allowing these actions to be discovered and executed through standardized APIs. End users can interact with these AI Actions using clear APIs, consistent naming conventions, and effective error handling.

For more information, see Configuring Model Context Protocol (MCP) Server and Using MCP Server to invoke AI actions.

Added Service Insights Agent Skills

New Feature

Service Insights Agent Skills are pre-defined snippets of deterministic logic that can be embedded into AI Actions. Administrators can incorporate the pre-built samples available within the AI Actions interface. This can be embedded into workflows to automate calculations and queries. While end users do not interact with Skills directly, the outcomes of AI Actions improve reliability and performance. Using these Skills, tasks such as data aggregation, metric calculations, and routine queries can execute more consistently, reducing errors and enhancing the overall user experience. By offering a set of sample Skills in this release, administrators can explore the concept and familiarize themselves with its potential. However, creating or editing new Skills is planned for future releases.

For more information, see Installing AI Agent Skills and Using Agent Skills.

Ability to update existing records using the SFM agent

Enhancement

This enhancement introduces support for Update Record type transactions within the Service Flow Manager (SFM) Agent. Administrators can now configure and execute edit actions on work orders, allowing specific fields to be updated automatically based on value mapping. Technicians using the Go app will see these changes reflected after invoking the AI action and performing an incremental sync. This ensures that updates are applied consistently and reduces manual data entry, leading to a smoother workflow.

Previously, the SFM Agent supported only the creation of new records. This expands the agent’s applicability across service workflows and enables record updates using the SFM Agent. For more information, see Configuring Service Flow Manager Agent and Configuring AI Actions.

Clickable record names in AI Action responses for easier navigation

Enhancement

AI action responses now display record names such as Work Orders, Installed Products, or Locations as clickable links. Users can now directly navigate to the corresponding record detail pages in either the Salesforce Web interface or the Go application, streamlining access and improving workflow efficiency.

Previously, AI action responses displayed record names as plain text, requiring users to manually search for the relevant records. With this enhancement, end users can quickly access record details directly from any AI Action response, including those launched from SFM wizards or via SFA automations. This improves navigation and usability across platforms and aligns with the existing behavior in AI Chat on Go. For more information, see Runtime Execution of AI Actions, Using AI Actions in Go, and Using AI Actions in SFM Wizard.

Clickable record names in AI chat responses for improved usability

Enhancement

AI chat responses in the Salesforce Web now display record names such as Work Orders, Installed Products, or Locations as clickable links. Selecting a link opens the corresponding record details page in Salesforce.

Previously, AI chat displayed record names as plain text. With this enhancement, end users can quickly access record details and take further action using SFM wizards. This improves navigation and usability within the Salesforce Web experience and acts as an advanced search engine to locate records in ServiceMax. For more information, see Using ServiceMax AI Chat in Go, Using ServiceMax AI Chat in Salesforce Web, and Using ServiceMax AI Bot in Zinc.

Expanded knowledge access sources in AI Action responses

Enhancement

AI action responses now display source document names such as manuals, guides, reference documents as clickable links when the Knowledge Access Agent is used. These links allow users to navigate directly to the source material when an AI action is invoked either via SFM wizards or SFA automations either in Salesforce Web or the Go app, depending on the platform used.

End users can verify that AI interpreted the document correctly or explore additional details directly from the source. This improves trust, transparency, and traceability in AI-generated responses. For more information, see Runtime Execution of AI Actions, Using AI Actions in Go, and Using AI Actions in SFM Wizard.

AI Action design improvements on the Salesforce Web and Go app for better user experience

Enhancement

The redesigned AI action user experience on the Salesforce Web and Go application enables end users to initiate an AI action and continue working on other tasks in Salesforce, while it runs in the background. This enhancement addresses a key usability challenge.

Previously, long-running AI actions would block the screen and prevent users from performing other tasks. The processing indicator and results of the AI action are displayed in a new window. These improvements enable users to multitask and stay productive while waiting for the AI action output. For more information, see Using ServiceMax AI Chat in Salesforce Web and Using AI Actions in Go.

Multi-Agent AI Actions

Enhancement

Admins can define AI Actions that coordinate multiple agents using natural language instructions. End users, such as technicians, can trigger these actions in various ServiceMax applications. ServiceMax AI automatically manages agent workflows by passing results between agents to retrieve documentation, check inventory, and create cases or work orders in one streamlined process. This reduces manual effort, speeds up field operations, and improves accuracy. For more information, see AI Actions.

Service Flow Manager (SFM) AI Agent

New Feature

The SFM AI Agent feature enables ServiceMax AI to execute SFM transactions based on admin instructions, enabling users to initiate actions such as creating follow-up Work Orders without navigating form-based UIs. This integration ensures AI-driven updates comply with Service Flow Manager’s validation and data mapping rules. Technicians can interact with SFM flows through the AI Actions feature, streamlining workflows and enhancing productivity. For more information, see Configuring Service Flow Manager Agent.

ServiceMax Core Agent Parts Queries

Enhancement

Technicians can now ask AI-powered questions about parts used or required on Work Orders, and parts availability in their trunk stock. This enhancement makes the ServiceMax Core Agent capable of answering parts-related queries out of the box, helping technicians prepare for jobs more effectively.

Examples of supported queries include:

This feature expands the scope of AI assistance for technicians by enabling quick access to parts-related information, reducing delays, and improving first-time fix rates.

For more information, see Using ServiceMax AI Assistant in Go.

Knowledge Access API

New Feature

The Knowledge Access API enables admins to upload and tag documents from any system, extending document ingestion beyond the default supported sources such as Salesforce Files, OneDrive, and SharePoint. This API-based configuration allows ServiceMax AI to access documents stored in custom content management systems, internal knowledge bases, or documents stored locally by admins. As part of this setup, admins can upload both PDF and plain text documents via API and apply metadata tags to improve retrieval accuracy. These documents become available to the Knowledge Access agent, enhancing its ability to surface relevant information during runtime and ensuring alignment with organizational knowledge assets. For more information, see Uploading Documents Using Knowledge Access API.

Enable AI Chat Toggle in Go

Enhancement

The Enable AI Chat Toggle within the Mobile Console introduces flexible access control for both admins and end users. Admins can enable or disable AI Chat at the profile or group level and copy configurations across multiple users for efficiency.

End users, such as technicians, can access AI Chat in the Go app only if their profile has the toggle enabled. If the toggle is disabled, the AI Chat interface is hidden from the user. However, users can still access AI Actions from supported interfaces.

By default, the toggle is off. Organizations must enable it to provide access to AI Chat, enabling for personalized experiences and a focused interface.

For more information, see Enabling AI Chat in the Go App.

Try AI Chat in AI Console

New Feature

The Try AI Chat feature enables admins to simulate AI chat interactions directly within the ServiceMax AI Console, without deploying to applications. This feature streamlines the testing of AI configurations such as AI Agents and AI Actions in a safe, non-production environment. It accelerates validation cycles and improves setup efficiency by replicating the end-user experience using either real or no-context records. For more information, see Using AI Chat in Console.

Cloning AI Actions

New Feature

The Cloning AI Actions feature enables admins to duplicate existing AI Actions in the AI Console. This helps reduce setup time when similar instruction patterns are required across objects or workflows. Admins can update the object and instruction text, validate the cloned copy, and save it without affecting the original AI Action. This ensures accuracy when reusing AI Actions across modules such as Go, SFM Wizard, and Service Board. For more information, see Cloning AI Actions.

Languages Supported

New Feature

AI chat now officially supports interactions in various regional languages, enhancing accessibility for non-English speaking users. The AI can reference documents written in one language, such as English, and provide answers in the preferred user language. For example, users can receive instructions on replacing a machine's battery in their regional language, even if the AI uses an English document to find the information. These new language capabilities improve the effectiveness of the Knowledge Access Agent, allowing admins to upload documents for translation and present content in the preferred user language. Overall, this enhancement aims to make AI chat more inclusive and user-friendly for a diverse audience. For more information about the list of supported languages, see Language Support and Translations.

Custom AI Actions

New Feature

Users can now create custom AI actions with specific instructions to perform specific jobs, enhancing the utility of AI for various workflows. These actions can summarize work orders, suggest parts needed for upcoming work, and summarize location service activity, providing users with the necessary information to be prepared before starting a repair job. Once customized, these actions perform jobs quickly and provide information in an easy-to-understand format, thereby saving users time searching for and analyzing data. For more information, see Configuring AI Actions and AI Actions Overview.

Automated AI Actions

New Feature

Automated AI Actions through Service Flow Automation (SFA) allows AI-driven actions, such as Summarize Work Order, to be triggered automatically based on SFA configuration during record creation or update. The output of these AI actions is stored in a designated field, such as a custom Work Order Summary field, making it easy for users to access and review the generated summaries. For more information, see Configuring Automated AI Actions in Service Flow Automation.

ServiceMax AI Analytics

New Feature

The ServiceMax AI Analytics dashboard provides administrators with detailed insights into the usage and engagement of ServiceMax AI features within their organization. This dashboard offers comprehensive data on user interactions, frequency of use, and agent deployment, allowing admins to see how people in their organization are utilizing ServiceMax AI in the field. It helps them monitor effectiveness, optimize AI tool usage, and make informed decisions for change management, business case building, and budgeting. By understanding who is using the AI tools and how they are being used, administrators can:

For more information, see Viewing ServiceMax AI Analytics.

Saved Prompts

New Feature

The AI Chat feature in the Go app allows users to save prompts they have used before, so they can reuse them for similar jobs without re-entering them. This boosts productivity by reducing repetitive input and offers streamlined, efficient user experience. For more information, see Saved Prompts.

DataGuide: Automatic Draft Creation from PDF/Image Forms

New Feature

Administrators can now generate initial drafts of digital DataGuide forms from PDF or image-based sources. This feature is especially useful for transforming scanned documents or unstructured digital forms into structured, editable formats. Using AI, the system analyzes the layout and content of the uploaded files to create a draft form that can be reviewed and customized for accuracy and personalization. This significantly reduces manual effort and accelerates the digitization process for users with DataGuide and AI licenses. For more information, see Creating a Form Using ServiceMax AI.

AI Actions in Service Board

New Feature

ServiceMax AI is now available for scheduling use cases within Service Board. Out of the box default capabilities include creation of AI actions that allow users to schedule Jobs by using generative AI. This feature is designed to minimize dispatcher effort in manual scheduling, and is not intended for route optimization. AI Actions can be combined with follow up options that may, for example, require users to identify nearby Jobs. Admins can also create custom AI Actions, such as Job summarization. To use this feature, a separate ServiceMax AI license is required.

For more information, see Using AI Actions in Service Board.

AI Chat Integration in Salesforce Web App

Enhancement

AI Chat is now accessible to all users, allowing them to have personalized chats and manage workflows more efficiently, like the Go app. Users can leverage AI Chat to enhance productivity in managing Work Orders, Locations, Installed Products, and Case workflows. For more information, see Using ServiceMax AI Assistant in Salesforce Web.

Knowledge Access Agent Supports Multiple Platforms

Enhancement

The Knowledge Access Agent can now access and collect documents stored on additional platforms like OneDrive and SharePoint. Previously, users could only access uploaded documents like technical manuals in Salesforce, which limited the agent's functionality to a single platform. This limitation has been removed, and you can integrate the agent with built-in document storage platforms. This enhancement allows administrators to manage documents and help users use AI Chat to retrieve information for the appropriate product. For more information, see Adding External Document Sources.

Record Links in AI Chat Responses

Enhancement

AI Chat in the Go app now includes clickable links for record names mentioned in responses. These links enable users to directly access and open the records within the Go app. This enhancement simplifies navigation and enhances the ability to access related records outside the scope of AI Chat or utilize the SFM wizard on those records. For more information, see Viewing Record Names and Knowledge Sources in Go Mobile App.

AI Chat Button for Follow-Up Questions in GO App

Enhancement

The Go app now features a follow-up chat button on AI Action responses. Previously, users could perform AI Actions like Summarize Work Order, but there was no direct way to ask follow-up questions about the AI-generated information. With this new button, users can easily transition from viewing AI-generated summaries to engaging in a conversation with the AI. This interaction allows users to obtain additional information or clarification about the current record, making it easier to complete their jobs efficiently. For more information, see Follow up Chat.

User Feedback Integration on AI Action Responses

Enhancement

Users can provide feedback on AI Action responses. Previously, this feature was only available for AI Chat messages, but now it is integrated into AI Action responses as well. Administrators use feedback on actions like Summarize Work Order to refine and rewrite instructions, providing better responses and ensuring that the system evolves to better meet user needs. For more information, see Providing Feedback.

AI Actions in SFM Wizards

Enhancement

AI Actions can now be launched directly from the Service Flow Manager (SFM) wizard within the Salesforce Web app. This enhancement allows users to initiate AI actions for various ServiceMax objects, including Work Orders, Cases, Locations, and Installed Products. By integrating AI capabilities directly into the SFM wizard, users can perform more comprehensive data analysis and insights across various aspects of their service operations. For more information, see Adding AI Action SFM Wizard Step.

Run AI Action in AI Console

Enhancement

Administrators can now evaluate AI Actions directly within the AI Console. Previously, administrators had to evaluate AI Actions in ServiceMax AI. With this update, they can run AI Actions against selected records directly in the AI Console. The Run AI Action button enables administrators to verify configuration changes without needing to use an app where AI Chat is enabled. For more information, see Viewing Run Details.