US 12,259,862 B2
Method and system for automatic management of critical computing metrics
Anitha Srinivasan, Plano, TX (US); Lane MacDougall, Keller, TX (US); Rob Locurto, Marlboro, NJ (US); and Dilan Weerasinghe, Bournemouth (GB)
Assigned to JPMORGAN CHASE BANK, N.A., New York, NY (US)
Filed by JPMorgan Chase Bank, N.A., New York, NY (US)
Filed on Aug. 4, 2023, as Appl. No. 18/230,463.
Prior Publication US 2025/0045259 A1, Feb. 6, 2025
Int. Cl. G06F 16/21 (2019.01); G06F 16/28 (2019.01)
CPC G06F 16/219 (2019.01) [G06F 16/288 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method for facilitating automated management of computing metrics, the method being implemented by at least one processor, the method comprising:
aggregating, by the at least one processor via an application programming interface, unstructured data from at least one computing component, the unstructured data including application log data;
parsing, by the at least one processor via a topic modeling algorithm comprising a statistical model algorithm with an unsupervised machine learning algorithm, semantics of the unstructured data to generate at least one structured data set, each of the at least one structured data set including textual data;
extracting, by the at least one processor, time series data from the at least one structured data set, the time series data relating to at least one performance metric of the at least one computing component;
generating, by the at least one processor using at least one model, a summary of the textual data in the at least one structured data set;
correlating, by the at least one processor using the at least one model, the summary of the textual data with the time series data;
determining, by the at least one processor using the at least one model, at least one characteristic for each of the at least one performance metric based on a result of the correlating, the at least one characteristic including a criticality characteristic;
detecting, by the at least one processor using the at least one model, at least one anomaly in the at least one performance metric;
determining, by the at least one processor using the at least one model, a first resolution action to remedy each of the at least one anomaly; and
initiating automatically, by the at least one processor using the at least one model, a self-healing action comprising the first resolution action based on historical resolution data and providing additional computing resources that automatically resolves factors contributing to predicted errors associated with the detected at least one anomaly.