Skill Recommendation components in Workforce Optimization for ITSM

  • Release version: Xanadu
  • Updated August 1, 2024
  • 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 Skill Recommendation components in Workforce Optimization for ITSM

    Workforce Optimization for ITSM enhances incident resolution by recommending relevant skills to agents. It leverages roles for access control, tables for storing skill data, configurable properties for tuning behavior, and a scheduled job to automate skill prediction and recommendation. This system uses both supervised and unsupervised machine learning models to identify the most suitable skills needed to resolve incidents efficiently.

    Show full answer Show less

    Key Features

    • Roles:
      • Skill Recommendation User (snsre.user): Allows viewing skill recommendation data.
      • Skill Recommendation Admin (snsre.admin): Grants administrative capabilities to configure skill recommendation properties.
    • Configurable Properties: Accessible via Skill Recommendation > Configuration, these properties allow administrators to:
      • Enable or disable skill recommendations.
      • Set the maximum number of skills predicted using supervised and unsupervised learning (default 3 each).
      • Define how many similar resolved incidents to analyze for predictions (default 15).
      • Specify the threshold for Predictive Intelligence to recommend a skill to an agent (default 20 predictions).
      • Choose or replace Predictive Intelligence solution definitions for supervised and unsupervised skill predictions tailored to incident resolution.
    • Scheduled Job: A daily job runs at 1 AM to analyze incidents closed the previous day, recommending skills that helped resolve those incidents for similar open cases.
    • Data Tables:
      • User Predicted Skill [snsreuserpredictedskill]: Tracks how often a skill is predicted for each user; entries not recommended in the last 60 days are automatically purged.
      • Task Predicted Skill [snsretaskpredictedskill]: Stores predicted skills associated with incident types; records older than 60 days are automatically deleted.

    Key Outcomes

    By implementing these skill recommendation components, ServiceNow customers can:

    • Improve incident resolution efficiency by assigning the most appropriate skills to agents based on historical data and machine learning models.
    • Customize and control the prediction parameters to align with organizational needs and incident types.
    • Maintain up-to-date skill recommendations through automated daily updates and data management.
    • Empower administrators and users with role-based access to monitor and manage skill recommendation processes securely.

    Workforce Optimization for ITSM uses roles to administer skill recommendation, tables to store skill data, and properties to modify default behavior, and scheduled job to recommend skills in configurable Workforce Optimization for ITSM.

    Roles

    Role title [name] Description Contains roles
    Skill Recommendation User [sn_sre.user] Grants rights to view skill recommendation tables. wfo.user
    Skill Recommendation Admin [sn_sre.admin] Grants administrative rights to edit the properties for skill recommendation.
    • wfo.admin
    • sn_sre.user

    Properties

    Navigate to Skill Recommendation > Configuration to configure these properties.

    Property Description

    Enable skill recommendation.

    sn_sre.enable_skill_recommendation
    Enable this property to start recommending skills for agents.
    • Type:true | false
    • Default value:true

    Maximum number of skills to predict based on supervised learning.

    sn_sre.max_supervised_skills
    Using supervised learning, the maximum number of skills to predict for each incident ordered by confidence of prediction.
    • Type:Integer
    • Default value:3

    Maximum number of skills to predict based on supervised learning.

    sn_sre.max_unsupervised_skills
    Using unsupervised learning, the maximum number of skills to predict for each incident ordered by confidence of prediction.
    • Type:Integer
    • Default value:3

    Number of resolved similar tasks to use for predicting skills.

    sn_sre.number_of_similar_incidents
    The number of resolved similar incidents to use to predict skills using supervised learning, ordered by confidence of prediction, to resolve similar types of incidents.
    • Type:Integer
    • Default value:15

    Number of times Predictive Intelligence must predict the same skill for an agent before recommending it for the agent.

    sn_sre.user_predicted_skill_threshold
    The number of times Predictive Intelligence must predict the same skill for an agent before recommending the skill for the agent.
    • Type:Integer
    • Default value:20

    Similarity solution definition to recommend skills from similar incidents.

    sn_sre.unsupervised_solution_definition_for_incidents
    Name of the Predictive Intelligence solution definition used for predicting skills to resolve incidents using unsupervised learning. If you have created your own solution definition, you can replace the default one with the one you have created.
    • Type:String
    • Default value:ml_sn_sn_sre_global_recommend_similar_skills_for_incidents

    Similarity solution definition to recommend skills for incidents.

    sn_sre.supervised_solution_definition_for_incidents
    Name of the Predictive Intelligence solution definition used for predicting skills to resolve incidents using supervised learning. If you have created your own solution definition, you can replace the default one with the one you have created.
    • Type:String
    • Default value:ml_sn_sn_sre_global_recommend_skills_from_similar_incidents

    Scheduled job

    Scheduled job Description
    Start skill prediction Runs the job every day at 1 AM on all incidents that were closed the previous day. Recommends the skills used to close the incidents to resolve similar open incidents.

    Tables

    Table Description
    User Predicted Skill [sn_sre_user_predicted_skill]
    • Stores how many times a skill has been predicted for the user.
    • Skills not recommended in the last 60 days are automatically deleted from this table.
    Task Predicted Skill [sn_sre_task_predicted_skill]
    • Stores the skills predicted to resolve each type of incident.
    • Tables created over 60 days ago are automatically deleted from this table.