Advanced Insight Module
This feature is part of the Advanced Insight Module. If your organization has not purchased this module, you may not have access to the feature.
If you are interested in upgrading to the Advanced Insight Module, contact your BigPanda account team.
Automated Incident Analysis speeds up incident triage and collaboration by giving your team plain-language incident titles and detailed descriptions built from enriched, actionable incident data.
Customize your Automated Incident Analysis settings from the AI Analysis Configuration screen. From this screen, you can select the alert tags used to generate the analysis, adjust instructions for the large language model (LLM) to use in generating summaries, and select the incident tags to use for similarity-based recommended actions.
Select which alert tags are used to generate AI Analyses.
Provide instructions to the Large Language Model (LLM) to customize AI summaries to fit your organization's workflows.
Determine which incident tags should be displayed for recommended actions.
Roles with the following permission can interact with the AI Analysis Configuration page in BigPanda Settings:
Permission Name | Description |
|---|---|
AI Module Configuration | View and modify AI module configuration. |
Permission access levels can be adjusted by selecting either View or Full Access. To learn more about how BigPanda's permissions work, see the Roles Management guide.
You can customize Automated Incident Analysis settings on the AI Analysis Configuration screen.
To edit the configuration:
Navigate to Settings > AI Analysis Configuration.
Select the Edit button.
In the AI Analysis configuration screen, navigate to the Alert Tags section and select the alert tags you would like to use to generate the AI analysis. You can type the tag name in the box, or select it from the drop-down menu.
In the AI Analysis section, populate the following fields:
Field
Description
Presets
Select a preset LLM instruction from the drop-down menu.
You can select from:
Localized Language Output - Translates the analysis to a local language. You can select from ISO 639-1 Standard Language Codes.
Technical SRE - Tailors the analysis for highly technical SRE teams.
Non-technical Audience - Structures the analysis for a non-technical audience. Uses clear, concise language with a focus on business impact.
Executive Summary - Provides a concise, high-level summary for an executive audience. Includes information about potential business and revenue impact.
Comprehensive Incident Analysis - Provides a comprehensive analysis and actionable insights tailored for effective incident resolution and system health improvement.
Network Engineer - Produces a detailed, technically focused analysis for an audience who has full knowledge of the network topology and devices.
Database Administrator - Tailors the analysis for a highly technical database administrator who has deep knowledge of database systems, configurations, and performance optimization
Adjust Instructions
The selected preset will appear in the text box. You can adjust the instructions for the LLM to fit your organization's workflows.
See the LLM Instruction Examples below for ideas.
If you need to revert your instructions to the preset, select Reset.
In the Recommended actions section, select up to 5 incident tags to use for similarity-based action recommendations.
Select a Similarity score threshold. If an incident has a similarity score equal to or above the configured threshold, it will be included in the recommended actions.
Click Save.
The instructions tell the LLM what adjustments should be made to the generated analysis. You can instruct the LLM to always include or exclude certain information, write the summary in a certain way, or other include other instructions that fit your organization's workflows.
LLM instructions should be written in natural language. Here are a few examples of instructions to update the prompt:
Return the analysis in Spanish (Latin American Spanish dialect).
Include potential impact and urgency in the title and summary.
Write the summary and title in a way that business and technical personas will understand. For example, when the check is 'synthetics monitor' you should display something more easily understood like 'URL Check'.
When providing your summary, root cause analysis, and reasoning, always highlight key components like impacted systems, system names, application names, service names, assignment group, and similar details. Additionally, include potential impact to other related systems and services. Make mention in the title and summary if it is impacting Production environments.
When providing your summary, root cause analysis, and reasoning, please always highlight the key components, system names, application names, service names, and similar details by using ALL CAPS.
Before saving your changes, you can generate a preview of the AI Analysis.
Select an incident ID from the drop down menus and click Generate to see how the summary will appear for a specific incident.
Learn more about accelerating incident management with the Advanced Insight Module.
Find details about viewing Automated Incident Analysis summaries in the BigPanda console.