Predictive Intelligence for CSM solution definitions

  • Release version: Yokohama
  • Updated January 30, 2025
  • 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 Predictive Intelligence for CSM solution definitions

    Predictive Intelligence for Customer Service Management (CSM) enables you to leverage your instance records to build tailored AI-driven solutions that enhance case handling and knowledge management. These solutions are accessible as templates on instances where both Predictive Intelligence and Customer Service Management are enabled. You can also create custom solution definitions to fine-tune AI behavior according to your business needs.

    Show full answer Show less

    Solution Definitions

    The solution definitions provide AI models for classification, clustering, and similarity to improve various aspects of case management:

    • Classification Solutions: Predict fields such as Assignment Group, Category, and Priority from the case’s Short description to automate case routing and prioritization.
    • Clustering Solution: Groups similar cases into topics based on Short description, helping identify common issues.
    • Similarity Solutions: Recommend similar cases (both open and resolved), major issues, and related knowledge articles by comparing case descriptions and knowledge article content. These recommendations assist agents in faster case investigation and resolution.
    • Auto-Responder Integration: The “Similar Knowledge Articles All” solution definition supports auto-responder notifications by automatically suggesting relevant knowledge content.

    Business Rules for Classification Solutions

    Specific business rules support the classification solutions (Assignment, Categorization, Prioritization), running on the Case [sncustomerservicecase] table. These rules automatically generate predictions when cases are created and update solution performance statistics when cases are closed.

    You can customize these business rules to align predictions and reporting with your operational requirements.

    Upgrade Guidance

    When upgrading from the Kingston to Yokohama release, create a new Default Case Based Prediction business rule using the provided template. This template includes methods for managing active solutions and supports customization for both global domain and domain-separated environments (e.g., MSPs).

    Maintaining Prediction Accuracy

    To ensure ongoing accuracy, manage prediction drift by retraining or modifying your solutions as business conditions evolve. Regular testing and adjustment of business rules across different user scenarios help maintain effective AI predictions and recommendations.

    Use your instance records to build Customer Service Management specific solutions.

    Solution definitions

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

    Table 1. Solution Definitions for Customer Service Management
    Solution Definition Solution Type Description
    CSM Case Assignment Classification Predicts the Assignment group field from the Short description.
    CSM Case Categorization Classification Predicts the Category field from the Short description.
    CSM Case Prioritization Classification Predicts the Priority field from the Short description.
    Grouping of Cases into Topics Clustering Clusters similar cases into topics based on the Short description.
    All Similar Cases Similarity Recommends similar cases based on the Short description that can help customer service agents with case investigation and resolution processes.
    Recommended Open Cases Similarity Recommends similar open cases based on the Short description.
    Recommended Resolved Cases Similarity Recommends similar resolved cases based on the Short description.
    Major Issue Detector Similarity Provides recommendations for major issues based on the Short description.
    • Recommends one or more major cases, if available.
    • Otherwise, recommends similar cases that are not linked as child cases to a major case.
    Similar Knowledge Articles Similarity Recommends similar knowledge articles by comparing the Text, Short description, and Description fields of knowledge articles to the Short description of the customer service case.
    Similar Knowledge Articles All Similarity Recommends similar knowledge articles by comparing fields in the Knowledge View database view [sn_customerservice_knowledge_view] with the Short description of the customer service case.
    Note:
    This solution definition is used in the Auto-Responder feature. For more information, see Train the similarity solution for finding Auto-Responder notification content.

    Business rules for classification solutions

    These business rules apply only to the CSM Case Assignment, CSM Case Categorization, and CSM Case Prioritization solution definitions. The rules are available only on instances where both Predictive Intelligence and Customer Service Management are active. Create your own business rules on the Case [sn_customerservice_case] table to customize prediction and reporting behaviors.

    Table 2. Business rules for Customer Service Management
    Business rule Table Description
    Default Case Based Prediction Case [sn_customerservice_case] Generates prediction results from the active Customer Service Management solutions. Runs when a case record is inserted.
    Update Prediction Results Case [sn_customerservice_case] Updates the solution precision and coverage statistics. Runs when a case record is closed.

    Upgrade Information

    If your instance is running on the Kingston release and you are upgrading to the Yokohama release:
    • Use the Default Case Based Prediction business rule template to create a new business rule. This rule includes a Solution variable that gets all active solutions retrieved by the findActiveSolution(solutionName) method.
    • In a global domain environment, use the solutionNames array variable which requires that you explicitly provide the solutions that are called by the business rule.
    • In a domain-separated environment, such as an MSP environment, refer to the commented code in the business rule template for easy customization.
    • The business rule template calls the applyPredictionForSolution() method to predict regardless of any changes to the default value.

    Maintaining prediction accuracy

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