Now Assist in Virtual Agent readiness on the ServiceNow AI Platform

  • Release version: Yokohama
  • Updated October 16, 2025
  • 3 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Now Assist in Virtual Agent readiness on the ServiceNow AI Platform

    Now Assist enhances ServiceNow's Virtual Agent by integrating AI-driven capabilities powered by large language models (LLMs). It enables natural language understanding without relying on traditional keyword matching or manual configurations. This results in virtual assistants that can guide users through complex tasks, provide high-confidence answers, and improve user engagement with conversational AI functionality.

    Show full answer Show less

    Key Features

    • AI Asset Discovery: Automatically discovers and matches user intents to Virtual Agent topics and other AI assets such as generative AI skills, agentic workflows, subflows, and actions, eliminating the need for time-consuming keyword or NLU configurations.
    • Simplified Deployment: LLM-powered Virtual Agent topics allow faster rollout and improved conversation quality, reducing manual tuning efforts.
    • AI Search Genius Results: Users receive curated, actionable responses through Genius cards, which summarize knowledge and provide direct actions like "Request this item."
    • Conversational Catalog Ordering: Users can order Service Catalog items using natural language conversations. The Virtual Agent asks clarifying questions and confirms intent before processing the request. (Note: Catalog items must be marked as conversational.)
    • Multi-turn Q&A: Supports seamless follow-up questions for refining queries and improving answer relevance.

    Implementation Guidance

    Setting up Now Assist requires configuring or creating an LLM assistant and assigning it to one or more portals. Without associating LLM Virtual Agent topics to an assistant, those topics will not be discoverable.

    A high-level readiness checklist includes:

    • Install Now Assist in Virtual Agent: Available via the Conversational Interfaces admin console, typically after installing a Now Assist product like Now Assist for ITSM.
    • Review Virtual Agent Topic Inventory: Identify high-volume user intents using Conversational Analytics dashboards and Automation Discovery reports to prioritize self-solve opportunities.
    • Review Knowledge Base: Align relevant KB articles with top user intents to simplify topic management and enhance self-service capabilities.
    • Migrate NLU Topics to LLM: Use the migration tool to convert existing NLU or keyword-based topics to LLM-powered topics, preserving prior investments while leveraging AI improvements.
    • Review Service Catalog Items: Identify self-serve catalog items that overlap with LLM topics to avoid duplication and reduce topic creation efforts.
    • Leverage Pre-built LLM Topics: Use Now Assist’s pre-built LLM topics as a foundation to minimize rework and accelerate deployment.

    Best Practices and Additional Considerations

    • Optimize topic descriptions during migration to clearly reflect user intent and expected outcomes.
    • Customize the assistant’s branding and chat experience through guided setup options, choosing between standard or enhanced chat interfaces.
    • Integrate Now Assist with Microsoft Teams for enhanced collaboration and accessibility.

    What to Expect

    By implementing Now Assist in Virtual Agent, ServiceNow customers can expect more intuitive, AI-powered conversational agents that reduce manual configuration, improve self-service success rates, and deliver actionable, natural language responses. This leads to faster deployments, better user engagement, and streamlined catalog ordering processes within Virtual Agent.

    Now Assist enhances Virtual Agent with AI-driven capabilities that understand natural language, guide users through complex tasks, and deliver high-confidence answers without relying on rigid keyword matching or manual configurations.

    Now Assist in Virtual Agent provides the following features:
    • AI asset discovery

      Say goodbye to time-consuming keyword or NLU configurations. Now Assist uses LLMs to automatically discover and match user intents to Virtual Agent topics and other AI assets, including generative AI skills, AI agents and agentic workflows, and subflows and actions.

    • Simplified deployment

      Using LLM-powered Virtual Agent topics, teams can accelerate rollout and improve conversation quality. This means less effort spent on manual tuning and more time delivering value.

    • AI Search Genius Results

      Users receive curated, actionable responses via Genius cards, which contain summarized knowledge with direct actions such as Request this item.

    • Conversational catalog ordering

      Users can request Service Catalog items using natural conversation. Virtual Agent asks clarifying questions and confirms the user's intent before completing the request.

      Note:
      Service Catalog items must be marked as conversational to work with Virtual Agent. For details, see Catalog item conversational details page overview.
    • Multi-turn Q&A

      Follow-up questions are handled seamlessly, allowing users to refine their queries and get better answers.

    Setting up Now Assist in Virtual Agent requires customizing or creating a new LLM assistant. You can assign an assistant to one or more portals. If LLM Virtual Agent topics aren't associated with an LLM assistant, they aren't discoverable.

    High-level checklist

    1. Install Now Assist in Virtual Agent
    You can install it from the Conversational Interfaces admin console once you have installed a Now Assist product such as Now Assist for IT Service Management (ITSM).
    To set up Now Assist in Virtual Agent, you configure an assistant.
    See: Configuring assistants overview
    2. Review your Virtual Agent topic inventory
    Review your topics and identify high-volume user intents. You can use the Conversational Analytics dashboard and Automation Discovery reports.
    Why? This helps you identify the top self-solve opportunities in Virtual Agent.
    See:
    Review your knowledge base
    Identify KB articles that can self-serve any of the top intents you identified.
    Why? This simplifies topic management and enables self-service.
    See: Knowledge Base readiness for Now Assist on the ServiceNow AI Platform
    4. Migrate NLU topics to LLM
    Use the topic migration tool in Virtual Agent to convert NLU topics to LLM.
    Why? Leverage existing Virtual Agent topics with minimal effort.
    See: Migrating NLU/keyword Virtual Agent topics to LLM topics
    5. Review Service Catalog items
    Identify self-serve catalog items that can be replaced with existing LLM topics.
    Why? To avoid redundancy and eliminate the need to create new Virtual Agent topics.
    See: Service Catalog readiness for Now Assist on the ServiceNow AI Platform
    Review LLM Virtual Agent topics that come with Now Assist
    Use these LLM topics as a starting point for Virtual Agent topic creation.
    Why? New LLM versions of older NLU Virtual Agent topics reduce rework.
    See: ITSM Virtual Agent pre-built LLM topics

    Tips

    • When migrating legacy NLU topics, ensure that you optimize topic descriptions so that the topic is clearly described and aligned with the intent and expected results.
    • You can customize the look of your assistant and the chat experience during guided setup.

      For details, see Brand an assistant.

    • You can choose the chat experience you want for each assistant:
    • You can integrate Now Assist in Virtual Agent with Microsoft Teams.

      For details, see Integrating Now Assist in Virtual Agent with Microsoft Teams.

    For more information about conversational catalogs in AI, see the following information from ServiceNow Community and YouTube: