Localized Incident Resolution Generator with Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current IT help desk systems are inefficient in resolving IT incidents due to slow resolution times, low first contact resolution rates, and the inability to provide customized, localized solutions, leading to increased storage requirements and processing times.
Innovation Solution
The system generates fresh, customized resolutions for IT incidents using retrieval transformers, reconstructs incidents in a safe environment, generates RPA workflows, determines the best candidate solutions, and deploys successful RPA workflows to resolve incidents in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If current automated systems store substantially all information regarding incidents in the base models, then comprehensive information is available, but storage requirements dramatically increase and processing times substantially increase
Solution Approach 1:
The system extracts only the relevant incident attributes needed for resolution from the complete incident information, storing them separately in an incident attributes database rather than storing all raw incident data in base models. This extraction approach maintains information completeness while reducing storage requirements and processing time.
Solution Approach 2:
The incident information is segmented into distinct attributes (incident type, severity, affected systems, resolution steps, etc.) that are stored in a structured database format. This segmentation allows the system to retrieve only specific attributes needed for each incident type, rather than processing all incident data, thereby reducing processing time while maintaining comprehensive information availability.
2Quantity of substance
If current automated systems provide generic information from historical resolutions, then storage requirements are reduced, but the solutions are cryptic and not adapted to current IT incidents
Solution Approach 1:
The system tailors resolution information to the specific local context of each incident by matching incident attributes with corresponding resolution attributes from historical data. Instead of providing generic resolutions, the system adapts the resolution information to match the specific characteristics of the current incident, making the solutions more usable and effective.
Solution Approach 2:
The system pre-processes and structures resolution information during incident creation, organizing it by incident type and attributes. This preliminary organization allows the system to quickly retrieve and present relevant, adapted resolution information without requiring extensive processing at the time of incident resolution, maintaining both efficiency and usability.
3Ease of operation
If human technicians provide initial first response to support tickets, then customized solutions can be provided, but first response times are slow and first contact resolution rates are low
Solution Approach 1:
The system enables self-service by automatically analyzing incident attributes, retrieving relevant resolution information from the database, and presenting customized resolution steps to users without requiring human technician intervention for initial response. This automation maintains solution customization while dramatically improving first contact resolution rates and reducing response times.
Solution Approach 2:
The system incorporates feedback mechanisms where resolution outcomes are captured and used to continuously improve the incident attributes database. This feedback loop ensures that the system learns from actual resolution successes and failures, enhancing the quality of automated resolutions over time and maintaining high customization without human intervention.
Data Source
AI summary
Localized incident resolution with reconstruction is disclosed. Fresh (i.e. non-generic) resolutions for IT incidents that include local user-specific attributes are generated. N-possible robotic process automation (“RPA”) workflows are generated for applications at issue or cosine similar ones. IT incidents are reconstructed in safe environments. Modeling is performed to determine which of the N-possible RPA workflows have a sufficient probability of resolving the IT incident. Select RPA workflows are executed on the reconstructed IT incident in the safe environment to identify which workflow(s), if any, resolve the IT incident. RPA workflows demonstrated to work can be localized based on the user-specific attributes and deployed automatically. Detailed feedback and instructions specific to the user and the user's IT incident can be provided to identify what steps to take if a complete automated solution was not identified and successfully tested.