Issue Resolution Automation Using Incident Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current IT landscape issue resolution procedures are inefficient due to the lack of capturing expertise in resolving issues, requiring repetitive analysis and manual searches for configuration parameters across multiple systems, leading to delayed automation and extensive human intervention.
Innovation Solution
An issue resolution automation system that processes incident reports, retrieves relevant features and parameters, and generates solutions using pattern-matching algorithms, allowing for automated implementation and optimization of configuration parameters to resolve issues efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual analysis and execution of processes by IT personnel is used to solve system errors, then expertise can be applied to identify solutions, but the expertise is difficult to capture or replicate and progress is delayed
Solution Approach 1:
The system creates a digital copy of IT personnel expertise by extracting features from resolved incidents and storing them in a knowledge base. This copy can be automatically retrieved and applied to similar future incidents, replicating expert knowledge without requiring the actual expert's time and effort.
Solution Approach 2:
The system performs preliminary actions by pre-processing incident data, extracting features, and storing resolved incident patterns in advance. When a new incident occurs, the system can quickly retrieve and apply pre-analyzed solutions, eliminating the need for real-time expert analysis and accelerating the resolution process.
2Extent of automation
If automated reconstruction of entire processes is attempted, then automation progress can be made, but input from operation team is needed which delays progress
Solution Approach 1:
The system extracts only the essential features and parameters needed for automation from the entire incident resolution process, rather than requiring complete process reconstruction. This selective extraction enables targeted automation of specific resolution steps without needing comprehensive input from operation teams about entire processes.
Solution Approach 2:
The system enables self-service automation by automatically retrieving relevant features from the knowledge base and generating resolution actions without requiring continuous human intervention. The automation system serves itself by using stored patterns to resolve incidents independently, reducing delays caused by waiting for operation team input.
3Reliability
If all configuration parameters are monitored and analyzed, then comprehensive issue detection is achieved, but data extraction and storage requirements increase significantly
Solution Approach 1:
Instead of uniformly monitoring all configuration parameters, the system applies local quality by identifying and monitoring only the specific features and parameters relevant to each type of incident. This selective monitoring approach maintains comprehensive issue detection for critical parameters while reducing overall data volume by ignoring irrelevant parameters.
Solution Approach 2:
The system segments the large set of configuration parameters into distinct feature categories based on incident types and resolution patterns. By organizing parameters into segmented groups and only extracting relevant segments for each incident, the system achieves comprehensive detection where needed while minimizing overall data extraction and storage requirements.
Data Source
AI summary
Implementations of the present disclosure include methods, systems, and computer-readable storage mediums for issue resolution based on actual use of configuration parameters. Actions include receiving, from a monitoring system, an incident report including a description of an issue of a process and a context of the issue, retrieving features associated with the issue based on the context of the issue, processing the features to extract a set of solutions that were executed to resolve associated issues, processing the set of solutions to generate a solution for the issue, comparing an accuracy of the solution with a solution implementation threshold, and implementing the solution to resolve the issue.