Based on the AI platform algorithm model, AISWare AIOps will focus on the three directions of quality assurance, cost management and efficiency improvement to conduct component-based packaging of intelligent solution capabilities for the entire process of fault discovery, diagnosis, self-recovery, and prediction, as well as scenarios such as intelligent decision-making, intelligent Q&A, capacity planning, resource optimization, etc., build reusable and evolvable operations learnware, and externally provide capabilities in the form of Open API. It supports the rapid implementation of intelligent operations demands, simplify the procedure, and reduce the cost of intelligence integration of the operations system.
Automatic matching with optimal algorithm based on features of historical data for anomaly detection
Locating the root cause of fault through intelligent mining of alarm RCA and link graphs
Realizing data trend prediction and early fault warning based on a variety of association algorithms
The product introduces AI technology and builds learnware capabilities with self-developed algorithm models for the needs of complex operations scenarios based on long-term accumulation. Compared with open source algorithms, it is more suitable for actual scenarios and produces better effects.
Providing componentized operations learnware capabilities, integrating with the operations system in the manner of Open API, simplifying the introduction method, so as to avoid duplication of construction, and reduce the cost of intelligence integration.
Realizing the support of intelligent operations capabilities through the combination of “platform + learnware” to provide integrated services of building, opening, operation and management of learnware.
Improving overall operations efficiency
System operation quality assurance
Reasonable control of operating costs
Enhancement of per capita operations capabilities
A telecom operator introduced AISWare AIOps for monitoring and alarm of multiple indicators such as Haproxy response delay, Kafka Topic traffic and business volume to achieve intelligent anomaly detection of dynamic threshold, which has successfully predicted several production faults.
A telecom operator introduced AISWare AIOps in the alarm convergence scenario of O-domain, which conducts real-time positioning of root causes for alarms and convergence through the dynamic mining of RCA rules of alarm, and effectively alleviates the alarm storm.