US 11,809,859 B2
Coordinated source code commits utilizing risk and error tolerance
Rafael de Souza Lima Espinha, Rio de Janeiro (BR); Priscila Vieira de Sousa, Campinas (BR); Silvana Bordini Coca Machado, Campinas (BR); and Marco Aurelio Stelmar Netto, Sao Paulo (BR)
Assigned to Kyndryl, Inc., New York, NY (US)
Filed by KYNDRYL, INC., New York, NY (US)
Filed on Mar. 25, 2021, as Appl. No. 17/212,497.
Prior Publication US 2022/0308862 A1, Sep. 29, 2022
Int. Cl. G06F 8/71 (2018.01); G06N 5/04 (2023.01)
CPC G06F 8/71 (2013.01) [G06N 5/04 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method, the method comprising:
receiving, by a processor of a computer device, data regarding a reference source code commit;
identifying, by an artificial intelligence model executed by the processor, a first group of source code commits including source code commits similar to the reference source code commit, wherein each source code commit in the first group is associated with a repository;
determining, by the artificial intelligence model, a first risk associated with implementing a first source code commit from the first group of source code commits;
determining, by the processor, an error budget associated with a first repository associated with the first source code commit; and
determining, by the processor, a first time for implementing the first source code commit based on the first risk and the error budget.