Machine Learning solutions for Flow Designer

  • Release version: Zurich
  • Updated January 20, 2026
  • 2 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 Machine Learning solutions for Flow Designer

    Machine Learning solutions for Flow Designer enable ServiceNow customers to integrate Predictive Intelligence models directly into their flows without scripting. This capability allows users to deploy machine learning solutions within their instances using a specialized spoke that provides ready-to-use actions for prediction tasks.

    Show full answer Show less

    The spoke depends on the Predictive Intelligence plugin and requires proper licensing. It supports different machine learning frameworks to predict, recommend, and organize data outcomes, enhancing automation and decision-making in workflows.

    Key Features

    • Frameworks provided:
      • Classification framework: Used for predicting categorical outcomes.
      • Similarity framework: Used for finding similar records based on defined criteria.
      • Clustering framework: Used for organizing data into groups.
    • Spoke actions: The spoke offers multiple actions that allow flows to interact with machine learning models, such as:
      • Classification Prediction (single and batch)
      • Similarity Prediction to find similar records
      • Regression Prediction (single and batch) – note that regression is deprecated in Zurich, so new regression models cannot be created, only existing ones used.
      • PI Confidence Check to compare prediction confidence values against user-specified thresholds
    • Requirements and dependencies:
      • A sharedservice.worker user is needed to train machine learning solutions.
      • Pre-trained Predictive Intelligence solutions must be available for use.
      • Dependent plugins include Predictive Intelligence and Predictive Intelligence Reporting plugins, which may require licenses.
    • User roles: The mladmin role grants full access to Predictive Intelligence features, allowing control and management of machine learning models and actions within Flow Designer.

    What This Enables for ServiceNow Customers

    By leveraging Predictive Intelligence for Flow Designer, customers can enhance automation workflows with machine learning predictions without needing to write custom scripts. This integration simplifies operational processes such as classification, similarity detection, and clustering, improving efficiency and decision accuracy. Customers can expect to incorporate predictive insights seamlessly into their business flows, enabling smarter, data-driven actions and improved service delivery.

    With Predictive Intelligence for Flow Designer (com.snc.ml_flowdesigner), you can deploy machine learning solutions in your instance. This spoke provides actions to incorporate Predictive Intelligence model predictions into flows.

    Predictive Intelligence activation

    This spoke requires Predictive Intelligence (com.glide.platform_ml). For more information, see .

    Key features

    Predictive Intelligence for Flow Designer provides three frameworks that you can use to create machine-learning solutions in your instance. Each framework delivers a different solution type for training the system to predict, recommend, and organize data outcomes.
    • Classification framework
    • Similarity framework
    • Clustering framework

    Spoke requirements

    • A sharedservice.worker user to train solutions.
    • A pre-trained Predictive Intelligence solution.

    Spoke dependencies

    If you’re having trouble installing the app, ensure that these dependent plugins are installed:
    • Predictive Intelligence (com.glide.platform_ml) plugin
    • Predictive Intelligence Reporting (com.glide.platform_ml_pa) plugin
    Note:
    Some plugins may be licensable features and require appropriate licenses, if used outside the spoke implementation.

    Spoke actions

    Predictive Intelligence for Flow Designer (com.snc.ml_flowdesigner) provides actions to make predictions using existing models without having to write or maintain script. Available actions include:

    Action Description
    Classification Batch Prediction Obtain a predicted value from an active classification solution definition using multiple input records.
    Classification Prediction Obtain a predicted value from an active classification solution definition using a single input record.
    PI Confidence Check Compare an output value (number) from Predictive intelligence with a number specified by the user. For example: Compare the confidence value of a prediction with a specified value.
    Regression Batch Prediction Obtain a predicted value from an active regression solution definition using multiple input records.
    Note:
    The regression framework is deprecated in the Zurich release. You can continue to use existing regression solutions but you can't create new ones.
    Regression Prediction Obtain a predicted value from an active regression solution definition using a single input record.
    Note:
    The regression framework is deprecated in the Zurich release. You can continue to use existing regression solutions but you can't create new ones.
    Similarity Prediction Obtain similar records that exist in the table specified by the user in their similarity solution definition.

    Spoke user roles

    Predictive Intelligence for Flow Designer provides these user roles to control access to data:
    User role Description
    ml_admin Grants access to all Predictive Intelligence features