ML Incident Prediction for Application Change Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software applications lack the ability to predict undesirable incident occurrences resulting from changes, leading to disruptions and financial/reputational damage in enterprises.
Innovation Solution
A Machine Learning (ML) model is trained to determine the probability of incidents from application change attributes, using historical data and NLP techniques, and integrated with IT management workflows to provide early detection and proactive remediation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If existing software applications are used for managing changes, then change management processes are automated, but the ability to predict incident occurrences is lost
Solution Approach 1:
The system performs preliminary analysis of change attributes before implementation by training an ML model on historical data. The model predicts incident probability in advance, allowing administrators to take preventive actions before changes are executed, thus resolving the contradiction between automation and prediction capability.
Solution Approach 2:
An ML prediction model is introduced as an intermediary component between change management processes and incident occurrence. This mediator analyzes change attributes and provides probability predictions, enabling automated change management while maintaining incident prediction capability through the intermediary analytical layer.
2Speed
If changes are implemented without prediction capabilities, then operational speed is maintained, but disruptions and financial impact increase
Solution Approach 1:
The system performs preliminary prediction of incident probability before change implementation. By knowing the risk probability in advance, administrators can proceed with confident, speed-efficient changes or adjust them proactively, preventing disruptions without slowing down legitimate change operations.
Solution Approach 2:
The system provides feedback on incident probability to administrators before changes are implemented. This feedback loop allows for rapid decision-making where changes can be executed quickly when safe, or adjusted/deferred when risk is high, thus maintaining operational speed while reducing harmful disruptions.
3Measurement precision
If historical data is processed to train ML models, then incident prediction accuracy is improved, but data processing complexity increases
Solution Approach 1:
The system extracts and isolates relevant features from historical data into distinct change attributes (e.g., change type, timing, scope). By extracting and separating these attributes, the complex historical data is transformed into structured input features that can be efficiently processed by the ML model, reducing processing complexity while maintaining prediction accuracy.
Solution Approach 2:
The data processing pipeline is segmented into distinct stages: data collection, feature extraction, model training, and prediction. This segmentation allows each stage to be optimized independently, managing overall complexity through modular processing while achieving high prediction accuracy through specialized handling of each data component.
Data Source
AI summary
A computer-implemented method and system for determining a probability of incident occurrence resulting from one or more changes to one or more computer networked applications. A machine learning (ML) model is trained to identify probability of likelihood of one or more incidents from occurring to one or more computer applications attributable to one or more changes to the one or more computer applications. Information is received corresponding to at least one application change attribute, which is then analyzed by the trained ML model to identify one or more incident occurrence indicators applicable to the at least one application change attribute. A probability of one or more incidents occurring is then determined corresponding to the at least one application change attribute.