AI Network Incident Analysis Using Historical RCA and Intent Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional root cause analysis (RCA) in network incidents is performed manually by operators, which is inefficient and lacks automation and accuracy.
Innovation Solution
A computer system utilizing AI models to process historical RCA documents, identify intents, and generate responses for network operators, including root cause analysis and recommended actions, using large language models and vector support machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual root cause analysis is performed by operators, then analysis can be conducted with human judgment, but productivity is low and manual effort is high
Solution Approach 1:
The system enables self-service automated root cause analysis where the network management system automatically detects incidents, queries historical data, identifies root causes, and generates RCA documents without requiring manual operator intervention for each analysis task
Solution Approach 2:
The patent replaces the mechanical manual analysis process with an automated system using large language models and vector support machines to perform root cause analysis, substituting human operators with AI-based computational mechanisms
2Reliability
If manual RCA documents are produced by operators, then detailed analysis can be captured, but device complexity increases due to manual processes
Solution Approach 1:
The patent replaces complex manual analysis processes with automated AI models including large language models for understanding incident descriptions and vector support machines for comparing incident vectors against historical data, reducing operational complexity while maintaining reliability
Solution Approach 2:
The system introduces an intermediary automated analysis layer between incident detection and RCA document generation, using AI models as mediators to process incident data, query historical records, and generate standardized RCA documents without direct manual intervention
3Productivity
If automated systems are implemented for RCA, then productivity increases and manual effort decreases, but measurement precision may be affected
Solution Approach 1:
The system uses large language models as intermediaries to accurately interpret incident descriptions and vector support machines as intermediaries to precisely compare incident vectors with historical data, maintaining measurement precision through specialized AI components designed for specific analytical tasks
Solution Approach 2:
The system performs preliminary actions by pre-processing incident data into standardized vectors, pre-training models on historical RCA documents, and pre-establishing comparison metrics before actual root cause analysis occurs, ensuring precision is built into the automated process from the start
Data Source
AI summary
Systems, devices, and methods related to network incident analysis. An example method includes: receiving historical RCA documents specific to a network service provider, receiving network data associated with the RCA documents, processing the RCA documents to generate RCA data based on the RCA documents and the network data, generating one or more vectors based on the RCA data and the network data, constructing and training one or more AI/ML models based on the RCA data and the vectors, receiving a query from a network operator of the network service provider, identifying one or more of the historical RCA documents pertaining to the query using the AI/ML models, analyzing the query using the AI/ML models to extract one or more intents of the network operator, generating contents using the AI/ML models, and generating a response comprising the contents for output.