HR Predictive Intelligence Workbench implementation

  • Release version: Yokohama
  • Updated January 30, 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 HR Predictive Intelligence Workbench implementation

    The HR Predictive Intelligence Workbench (PIWB) enables ServiceNow customers to leverage machine learning (ML) to enhance HR service workflows. By training and implementing predictive ML models, customers can optimize business processes such as case categorization, assignment prediction, knowledge recommendations, and resolution time forecasting. This solution supports both guided and classic templates for rapid setup and customization.

    Show full answer Show less

    Key Features

    • Use Case Templates: Customers with the snpiwbhrcontent.admin role can select from pre-built templates to create ML models. Templates are classified as:
      • Guided Templates: Provide step-by-step setup assistance and include auto-trained models for faster deployment.
      • Classic Templates: Offer detailed configuration guidance to customize and train models based on specific business needs.
    • Solution Definitions: Templates available only when both Predictive Intelligence and HR modules are active, including:
      • Classification models for predicting HR service and assignment groups for incoming cases.
      • Similarity models that recommend relevant knowledge articles, similar cases, and identify knowledge gaps.
      • Regression models to estimate case resolution times.
      • Clustering models to detect HR case clusters lacking knowledge articles for targeted improvements.
    • Use Case Creation Phases: Includes creating and training models, testing predictions through single or batch testing, tuning model parameters, and integrating the best performing model into production workflows.
    • Prediction Accuracy Maintenance: Customers can manage model drift by retraining or modifying models to keep predictions aligned with evolving business conditions.

    Key Outcomes

    • Automated Email Case Categorization: Reduces manual effort and improves productivity by auto-classifying email cases.
    • Improved Case Routing: Predicts the appropriate HR service or assignment group to expedite case handling.
    • Enhanced Knowledge Management: Identifies relevant knowledge articles for agents and employees, discovers knowledge gaps, and supports knowledge base improvements driven by demand.
    • Faster Case Resolution: Provides estimated resolution times and recommends relevant resources to accelerate case closure.
    • Personalized Content Recommendations: Suggests articles and catalog items based on employee profiles for a tailored experience.

    Practical Implementation Notes

    • Use guided templates with auto-trained models to quickly move from setup to evaluation and integration.
    • Classic templates require more hands-on training and tuning but offer greater customization.
    • Administrators can create custom solution definitions to tailor predictive behavior beyond the provided templates.
    • Regularly monitor and maintain prediction accuracy by testing and updating models to reflect changes in HR operations.

    You can use machine learning to optimize your business processes. You can train and implement HR Predictive Intelligence Workbench use cases to augment your existing application workflows.

    Explore common use case templates

    With the sn_piwb_hr_content.admin role, you can explore the use case templates and create predictive machine learning models. To create a machine learning model, you first select a pre-built use case template. You can use one of the following templates for setting up the use cases.
    • Guided templates include a comprehensive setup process to help you through implementation. Templates with available auto-trained models accelerate your setup process, by providing a pre-generated model based on your data.
    • Classic templates include a comprehensive setup information to help you through implementation. Leverage the existing templates to configure, test, and train the models based on your business requirements.

    When a template indicates Auto-trained model available, this means you can go directly to the evaluation phase of the use case setup. If the auto-trained model is acceptable, you can directly integrate the model with your business processes. Otherwise, you can tune this model or create another model. You may change the name and description of the use case later.

    Solution definitions

    These solution definitions are available as templates on instances where both Predictive Intelligence and HR are active. Create your own solution definition records to customize the behavior.

    Table 1. HR Solution Definitions
    Solution Type Solution Definition Description Implementation
    Classification Predict the HR service for incoming cases Predicts the correct HR service for cases. Guided
    Classification Predict the assignment group for incoming cases Predicts the correct assignment group for cases. Guided
    Classification Email Case Categorization Auto-categorizes the HR service for the email cases for improved productivity by saving time and costs. Guided
    Similarity Similar Closed HR Cases Recommends similar cases closed in the past to assist an HR agent for faster and better resolution. Classic
    Similarity User profile based recommendation Recommends top 3 relevant articles and catalog items based on users with a similar profile for content discovery and personalized experience. Classic
    Similarity Similar HR Cases and knowledge Automates the discovery of knowledge gaps in your knowledge bases and recommends insights into improving knowledge that is driven by demand. Classic
    Similarity Similar Knowledge Articles for HR Task Displays related articles to assist employees in completing the HR or Content or Campaign to-dos. Classic
    Similarity Similar Knowledge Articles for HR Case Uses ML to identify the most relevant knowledge articles to assist an HR agent for faster and better resolution. Classic
    Regression HR Case Resolution Time Determines the resolution time expected a case by analyzing similar closed cases in past for better visibility and transparency. Classic
    Clustering Demand Insights: HR Case Clusters Need Knowledge Identifies the case clusters that do not have knowledge and helps with filling the knowledge gaps in your knowledge base. Classic

    Use case creation phases

    Creating a predictive machine learning model involves several phases. After you create and train your model, evaluate and tune, test prediction results, and integrate the model with your business process. Use case model creation phases include:
    • Create and train models: Define parameters to create a model that you train based on your unique data. Create multiple models as you tune and refine your models by defining the right combination of coverage and precision to use.
    • Test your models: Get prediction results from your models to decide which one is best to integrate with your business process. To see if a model returns a correct result, you can use either the single or batch testing process.
    • Integrate the best model: Deploy the best model into your business process. After you determine which model returns the best, correct result, integrate it into production.

    Prediction accuracy maintenance

    You can manage prediction drift by retraining, modifying, or creating new solutions to reflect changes in your business conditions. Test and modify your business rule over time to ensure it works as desired across multiple consumption points and user persona.