Configure AI Incident Prevention
The AI Incident Prevention module enables you to assess and analyze the risk associated with proposed changes in your IT environment. It calculates a risk score based on specific change fields and compares the planned change with historical data to identify potential risks.
Within the AI Incident Prevention web app, you can view your dashboards and configure all settings related to change risk.
The following sections are available in the web app for AI Incident Prevention users:
Dashboards
Configuration
Guides and Resources
Access Controls
Configuration modules
AI Incident Prevention dashboards and configuration modules are accessed through the Web App.
If your organization has both AI Incident Prevention and AI Incident Assistant, the following configuration modules will appear alongside other configuration modules in the Biggy web app. If not, you will only see the modules and dashboards available for AI Incident Prevention.
Dashboards
The Dashboards module gives you access to analytics and data related to your organization's change requests. Use the Change Analytics and Change Risk Prediction dashboards to find risk prediction information and help prevent incidents before they happen.
These dashboards are only available to organizations who purchased AI Incident Prevention.
See the AI Incident Prevention Dashboards documentation for more information.
Configuration
The Configuration module allows you to customize AI Incident Prevention to meet your organization's workflows.
In this module, you can view and customize configurations related to:
General Settings - Set up preferences related to the LLM.
Integrations - Connect the tools that AI Incident Prevention can access information from in real-time.
Change Risk - Configure settings related to Change Risk dashboards.
Manage Data - Upload documents and see an overview of your indexed documents across different source categories.
General Settings
In the General Settings section, configure LLM settings.
Under LLM Settings, configure settings related to your large language model provider. Updating this section requires approval from BigPanda.
Integrations
In the Integrations section, you can configure tools to connect with AI Incident Prevention. The following integrations are available:
Change Risk
In the Change Risk settings, configure risk evaluation criteria for your organization, adjust the weightings of each risk category, configure the Change Risk Dashboard settings, and determine if risk assessments will be passed to your ITSM system.
Org-Specific Risk Evaluation
In addition to the risk analysis AI Incident Prevention performs, you can also define your own evaluation criteria.
Under Evaluation Instructions, define additional evaluation guidelines for the LLM to follow.
Writing custom risk evaluation instructions
Write evaluation instructions the same way you would explain to a coworker what makes a change risky or not.
You can write these instructions in a free-form way or use if/else decision logic format.
Risk Category Weightings
Under Risk Category Weightings, adjust the weight of risk categories. These weightings determine how the final risk score is calculated.
Adjust the weightings using the sliding bar or by typing a number into the box on the right side of each category.
Total weightings
The total of all risk category weightings must equal 100.
The following risk categories are available:
Risk Category | Description |
|---|---|
Implementation Risk | Documented change plan quality, implementation complexity, and the nature of the change. |
Topological Impact Risk | Which parts of the environment will be affected and how critical they are. |
Team and Individual Risk | The calculated experience level and reliability of the team or individual performing the change. |
Historical Incident Risk | Likelihood of causing an incident based on historically similar changes that caused incidents. |
Organization-Specific Risk | Your organization-specific risk evaluation guidelines or criteria, defined in Evaluation Instructions. |
Dashboard Settings
Under Dashboard Settings, configure how the Change Risk Dashboard appears for users in your organization.
In the Default Time Range section, set the default time range for the Change Risk Dashboard. This determines which changes are displayed when users first visit the dashboard.
In the Data Visibility section, determine what information is displayed on change risk assessment pages. Use the Hide Individual Names toggle to determine whether or not names will appear in the Team or Individual Risk details tab.
Notification Settings
Under Notification Settings, configure notifications for high-risk changes.
Use the Enable Risk Notifications toggle to determine if notifications will be sent when a change exceeds the risk threshold.
When risk notifications are enabled, you can set the Risk Threshold for when notifications will be sent. Changes with a risk score at or above the threshold will trigger a notification.
Notification Destinations
The Notification Destinations section appears if you've enabled risk notifications.
Toggle Send to Channels on to send high-risk changes to Slack or Teams channels. Use the drop-down menu to select one or more channels to send changes to.
Toggle Send as Email on to send changes to specific email addresses. Enter a comma-separated list of email addresses to receive notifications in the text box.
Post Results to ITSM Record
If ServiceNow is your change tool, you can define where and how analysis outputs are written into change records.
Toggle Enable Work Note Publish on to include analysis fields from Biggy in ServiceNow change record work notes.
In the Risk Threshold section, define the risk score that a change must meet or exceed to be posted to a work note. If the score crosses the threshold again later, a concise update will be appended instead of the full preview.
Click the Add Analysis Fields drop-down to select the fields to be included. The following options are available:
Field | Description |
|---|---|
Overall Risk | Overall risk score and rating for the change. |
Risk Prediction | Analysis of potential risks and outcomes. |
Risk Reasoning | AI-generated reasoning explaining the risk assessment rationale. |
Suggested Mitigations | Recommended actions to reduce identified risks. |
After selecting analysis fields, the ServiceNow Work Note Preview appears.
Manage Data
The Manage Data page provides an overview of your indexed documents across different source categories. AI Incident Prevention uses indexed document data to accurately determine change risk.
See the Manage Data documentation for more information.
Guides and Resources
The Guides and Resources module of the Web App provides helpful information and tips for using AI Incident Prevention and information about the latest updates to the app.
The following sections are available in the Guides and Resources module:
Quickstart - Introduction to AI Incident Prevention, including best practices and examples.
Release Notes - Details about the latest AI Incident Prevention releases.
Access Controls
In the Access Controls section, you can create and manage users, control which tasks they can perform, and which areas of the web app they have access to.
AI Incident Prevention uses the same access control functionality as the AI Incident Assistant. See the Manage BigPanda Biggy AI Access Controls documentation for more information.