Incident creation agent

The Incident creation agent enables users to report issues using natural language and automatically creates incidents with the relevant details, ensuring faster and more accurate incident logging.

To enable and configure the Incident creation agent, do the following:

1.Log in to Neurons for ITSM as an Administrator.

2.Open the Configuration console and select AI Configuration Hub > Agentic AI > Agent Settings.

3.Enable the Incident creation agent using the toggle button.

An additional toggle button with the same function is available in the upper-right corner of the agent page.

4.Click Configure to further set up the agent.

5.Under Knowledge base settings > Suggest articles before ticket creation, enable the toggle button if you want the agent to suggest knowledge articles before creating an incident.

6.Under Knowledge base settings > User can skip viewing knowledge base, enable the toggle button to provide users an option to skip the knowledge article search and go directly to Incident creation.

7.In Skip button name, enter a custom name for the skip button according to your preference.

8.Under Messaging > Successful creation, enter a success message that the agent shows after the successful Incident creation.

9.Under Incident handling, configure the following settings:

a. Default incident template:

Set the default incident template from the dropdown. When the agent cannot find a specific incident template for a user's query, it will use the default incident template you have configured.

In the template, do not leave any fields blank that cannot be answered by either the agent or the user. Ensure that the Field Value is filled for such fields.

For example, Team and Owner fields cannot be predicted by the agent, as the agent does not know to which team or owner the incident record will be assigned. In such cases, the agent might ask the user for the owner's details, but the user might not know them either.

Ensure the fields added in the template are available in the Default published incident form, otherwise the field will not appear in the Incident record.

To view the available templates, go to Incident Business Object > Quick Actions > Edit Actions > Graphical Action Designer / Classic Action Designer > Your Actions > Templates.

You can also create new templates using the Create New Actions option.

For more information, refer to Action Templates.

b. Default published incident form:

Select the appropriate incident form from the dropdown. The form helps understand the fields required for Incident creation when a user opts for Incident creation. Therefore, ensure that all the required fields are present in the Form for Incident creation.

If any required fields are missing in the form, you can create them from the Incident Business Object > Fields tab and then update the form. To update the Form, go to Incident Business Object > Forms tab > open the relevant form.

For more information, refer to Using Forms.

c. CI mining question depth:

Enter the maximum number of questions the Incident agent can ask to identify a specific user device.

d. Incident description question depth:

Enter the maximum number of questions the Incident agent can ask to better understand the user’s problem.

e. Device mining:

Enable this toggle to allow the Incident agent to search for possible devices that users might be experiencing issues with. If disabled, the agent does not ask about the device and creates the incident without device information.

If the agent cannot find the device through device mining, it prompts the user to manually provide the device name.

To ensure successful device mining, configure the following:

•Maintain device records in the CI

•Define the Employee - CI relationships

•Agentic configuration for Employee - CI relationships

•Enable the Full-text indexed option for relevant CI fields

  1. Maintain device records in the CI

    For agents to identify devices managed within your organization, ensure that device information is configured in Configuration Items (CI). The CI Business Object helps maintain a reliable record of the devices, enabling easy tracking agents.

    For more information, refer to Working with Configuration Item Management.

  2. Define the Employee – CI relationships

    To configure an Employee – CI relationship, open the Employee Business Object > Relationships tab.

    For the agent to identify the devices linked to a user/employee, you should configure the relationships between devices managed in CI records and associated users in your organization.

    The relationships configured here are used by the agent for device mining.

    For more information, refer to Using Relationships.

  3. Agent configuration for Employee – CI relationships

    For the agents to understand the relationship between an employee and CI, you must configure the relationship information in the Agentic configuration workspace as well.

    To configure a relationship:

    1. Log in to Neurons for ITSM as an Administrator.

    2. Open the Agentic Configuration workspace.

    3. Click NEW IVNT_AGENTICCONFIGURATION.

    4. Enter the following details:

    •Created By: The name of the admin who created the record.

    •Created On: The date and time when the record was created. Use the calendar icon to select the value.

    •Modified By: The name of the user who last modified the record.

    •Modified On: The date and time of the most recent modification. Use the calendar icon to select the value.

    •Configuration: Configure the required relationship names by referring to Employee record > Relationship tab.

    The following image shows the Agent Configuration page.

  4. Enable the Full-text indexed option for relevant CI fields

    To enable full-text search on relevant fields, go to CI Business Object > Fields tab, and enable Full‑text indexed option.

    The CI Business Object contains multiple device‑related fields, such as model number, model name, and serial number. For effective device identification, it is important to anticipate the type of device information users are likely to provide when prompted by an agent. To allow agents to search devices using this information, enable the Full‑text indexed option for corresponding fields.

    By default, some fields are already enabled for full‑text indexing.

    For more information, refer to About the Different Search Types.

f. Allow transfer to live agent: Enable the toggle button to transfer the interaction to your Service Desk Human Agent in case the Agentic AI bot cannot solve the user's request.

Along with this toggle, you must also enable the Allow transfer to live agent toggle in Global Settings > Escalation Handling.

10. Under Advanced settings > Incident description, define the scope of the Incident agent by detailing the types of incident queries the agent can handle. You can also provide some example queries related to incidents to help the agent better understand the context. When users ask queries other than those defined here, the agent considers them out of context and replies to users accordingly.

Modifying the intent description might result in unpredictable agent behavior. We recommend that only users with experience in LLM prompt engineering make these changes.

The following image shows the configuration page of Incident creation agent.

Fields marked with an asterisk (*) are mandatory.

Incident Analytics

To view the performance metrics and usage statistics for your Incident creation agent, open the Configuration console and select AI Configuration Hub > Agentic AI > Agent Settings > Incident creation agent > Configure > Analytics.

The Analytics page provides the performance of the Incident creation agent. It helps you monitor the following metrics:

•Number of Incident Agent Interactions: The total number of times users interacted with the Incident creation agent during Incident creation.

•Number of Incidents deflected through KS: The number of incidents successfully resolved by the AI using Knowledge Search, without creating an incident.

•Percentage of Incidents deflected through KS: The percentage of incidents resolved by the AI using Knowledge Search, calculated against total incident agent interactions.

•Number of Incidents created by IA: The total number of incidents created by the Incident Agent.

•Number of Transfers to Live Agents: The number of times the Incident creation process was transferred to a live agent.

•Percentage of incident transferred to Live Agents: The percentage of Incident creation interactions that resulted in a transfer to a live agent.

•Incident Completion Rate: The percentage of incident agent interactions that were successfully completed, based on 100 incident agent interactions.

The Time Range filter allows users to specify the period for which analytics data is displayed. Metrics on the Analytics page are updated based on the selected time range.