US 11,720,810 B2
Reducing mean time to find problems in enterprise information technology systems using bots
Rahul Chenny, Bangalore (IN); Ramshanker Kowta, Bangalore (IN); and Awadesh Tiwari, Bangalore (IN)
Assigned to Kyndryl, Inc., New York, NY (US)
Filed by Kyndryl, Inc., New York, NY (US)
Filed on Aug. 21, 2018, as Appl. No. 16/106,053.
Prior Publication US 2020/0065685 A1, Feb. 27, 2020
Int. Cl. G06N 5/04 (2023.01); G06N 20/00 (2019.01); G06F 16/28 (2019.01); G06F 16/23 (2019.01); G06N 5/043 (2023.01)
CPC G06N 5/043 (2013.01) [G06F 16/2365 (2019.01); G06F 16/283 (2019.01); G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method for leveraging Bots across various layers of an enterprise information technology system for reducing mean time to find problems (MTFP), the method comprising:
determining, by one or more processors, if one or more system Bots can identify one or more issues in an enterprise information technology system;
responsive to determining that the one or more system Bots cannot identify the one or more issues, escalating, by the one or more processors, the one or more issues to one or more process Bots;
invoking, by the one or more processors, one or more MTFP computation engines from related Bots in communication with the one or more process Bots;
identifying, by the one or more processors, the one or more issues in the enterprise information technology system, and further identifying at least one relevant related Bot of the related Bots, by the invoking of the one or more MTFP computation engines from the related Bots in communication with the one or more process Bots, and commanding the at least one relevant related Bot to determine at least one resolution to the one or more issue in the enterprise information technology system, wherein the at least one relevant related Bot has been determined to be relevant to the identifying the one or more issues in the enterprise information technology system;
updating, by the one or more processors, a knowledge repository with attributes of the identified one or more issues based on attribute data of the least one relevant related Bot, wherein the one or more process Bots cognitively learn from data stored on the knowledge repository;
outputting, by a user interface, the one or more identified issues to a user; and performing, by the one or more process Bots, a process action, wherein the process action runs an independent model to identify which Bot of the related Bots can identify a certain issue per a service level agreement (SLA) and initiate a process of assigning an identification task for identifying the certain issue to the identified Bot of the related Bots, wherein the one or more system Bots continuously tracks its internal system components for its health status, wherein the method includes determining that memory usage data is available to the one or more process Bots, wherein the memory usage data includes inconsistent garbage collection, high CPU temperature, and/or missing data, and responsively to determining that memory usage data is available to the one or more process Bots performing one or more decision optimizations to dynamically formulate templates to features of a detected and/or reported issue, wherein the features of the detected and/or reported issue include (i) application speed, (ii) probability of shut down, (iii) incorrect computations, (iv) transaction speed.