Alert Inheritance via ML Pattern Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
IT administrators often lack sufficient information to efficiently address IT alerts, leading to inefficiencies in managing and remediating IT system incidents.
Innovation Solution
The automatic inheritance of alert properties from similar resolved alerts using machine learning, which identifies common properties and updates the new alert with relevant metadata to prioritize and route alerts effectively, enabling preemptive action before incidents occur.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If IT administrators manually analyze and process alert information, then they can make informed decisions, but the time required to respond to and remediate alerts increases
Solution Approach 1:
The system performs preliminary analysis by automatically identifying similar resolved alerts and extracting their common properties before the IT administrator needs to make a decision. This pre-processing of information includes using machine learning to find patterns in historical alerts and pre-computing relevant metadata, so that when an alert occurs, the administrator receives pre-analyzed information immediately rather than having to manually search and analyze data.
2Measurement precision
If more information is provided to IT administrators about alerts, then diagnosis accuracy improves, but the complexity of the alert management system increases
Solution Approach 1:
The system extracts only the most relevant properties and metadata from similar resolved alerts, rather than presenting all available information. It identifies and extracts common properties such as alert patterns, affected components, and resolution methods from historical data, filtering out unnecessary information. This extraction approach provides targeted diagnostic information without overwhelming the administrator with excessive data or requiring complex manual processing.
3Measurement precision
If manual analysis of alert properties is performed, then accurate categorization is achieved, but productivity and efficiency decrease
Solution Approach 1:
The system performs self-service by automatically analyzing alert properties, identifying similar historical alerts, and categorizing new alerts based on patterns learned from resolved alerts. The machine learning model autonomously processes alert data, extracts relevant features, and assigns categories without requiring manual intervention. This automation maintains accurate categorization while significantly increasing processing throughput and reducing the time administrators need to spend on each alert.
Data Source
AI summary
A new alert is received. Machine learning is used to identify a plurality of resolved alerts similar to the new alert. One or more processors are used to automatically identify among properties of the identified similar resolved alerts, one or more common properties of the identified similar resolved alerts having one or more statistical metrics meeting one or more corresponding thresholds. The new alert is caused to inherit the identified one or more common properties.