Prompt Repository for AI Powered ITSM

The Prompt Repository is a curated collection of ready-to-use natural language prompts designed to guide users in effectively interacting with our AI-powered ITSM platform. These prompts span various functional domains such as Incident Management, Service Requests, Change Management, Problem Management, and CMDB (Configuration Management Database). Each prompt is carefully crafted to extract maximum value from AI features, ensuring users receive consistent, context-aware, and actionable responses from the virtual support agents and AI components.

The Prompt Repository acts as a bridge between the user and AI, significantly lowering the learning curve for both IT agents and end users. Instead of creating complex queries, users can use predefined prompts to perform tasks such as retrieving relevant incidents, analyzing change success rates, clustering related problems, or identifying missing CI relationships. This not only accelerates task resolution but also ensures a standardized approach to interacting with intelligent automation within the ITSM ecosystem.

For customers, the Prompt Repository improves efficiency, usability, and user confidence. It empowers users—regardless of their technical expertise—to leverage AI without the need for scripting or specialized training. By using intelligent prompts, organizations can drive faster adoption, improve agent productivity, reduce ticket backlogs, and enhance service quality.

Currently, we deliver the Prompt Repository during the implementation and training phase as part of our customer onboarding and enablement process. The initial set of prompts is customized based on the customer’s key modules and workflows. Our solution consultants work closely with customers to update and extend these prompts over time, enabling continuous alignment with evolving business and IT needs.

Best Practices for using prompt Repository

• Start with guided prompts before exploring free-form queries to become familiar with how the AI interprets instructions.

• Use module-specific prefixing when applicable (e.g., _Apex_KB_Analysis: for KB clustering tasks).

• Keep prompts clear and task-oriented, using keywords related to symptoms, categories, or desired outcomes.

• Avoid overloading prompts with multiple intents. Stick to single-focus instructions for more accurate results.

• Update or refine prompts based on user feedback and evolving workflows to maintain effectiveness.

Sample Prompts (End-User Persona)

Incident Management

• Show all open incidents assigned to me that mention Outlook crashing.

• Summarize the top recurring incident types reported in the last 7 days.

Service Request

• Track the approval status of my laptop replacement request.

• List all pending service requests related to software installation.

Change Management

• What is the risk level of the change request CR123456?

• Analyze recent successful changes related to the network team.

Problem Management

• Cluster all problems related to printer performance issues.

• What is the status of root cause analysis for PBI100234?

CMDB

• Find all CIs with missing relationships in the database cluster group.

• List the configuration items tagged as 'High Risk' and 'Outdated'.

Note

The execution and relevance of sample prompts are subject to the availability of integrated data sources, the specific ITSM modules configured in your environment, and the user's role-based access permissions. For optimal results, ensure that your environment includes the necessary data coverage and that users have the appropriate privileges.