US 11,900,275 B2
Proactively detecting and predicting potential breakage or support issues for impending code changes
Sossena Negussie, McLean, VA (US); Michael Mossoba, Great Falls, VA (US); and Joshua Edwards, Philadelphia, PA (US)
Assigned to Capital One Services, LLC, McLean, VA (US)
Filed by Capital One Services, LLC, McLean, VA (US)
Filed on Sep. 21, 2021, as Appl. No. 17/448,238.
Prior Publication US 2023/0088784 A1, Mar. 23, 2023
Int. Cl. G06F 8/65 (2018.01); G06N 5/04 (2023.01); G06F 8/77 (2018.01); G06N 5/022 (2023.01); G06F 11/36 (2006.01)
CPC G06N 5/04 (2013.01) [G06F 8/77 (2013.01); G06N 5/022 (2013.01); G06F 8/65 (2013.01); G06F 11/3664 (2013.01); G06F 11/3688 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for predicting code quality prior to deployment, the system comprising:
one or more memories; and
one or more processors of a device that includes computing hardware used in a cloud computing environment, coupled to the one or more memories, configured to:
obtain, from a project management tool, data related to an impending code change;
extract, based on the obtained data and based on performing natural language processing, one or more feature sets related to the impending code change,
wherein the one or more feature sets include one or more features that relate to an impact that the impending code change has on an existing code base and one or more features that relate to a historical code quality for a developer associated with the impending code change or a quality of a development session associated with the impending code change,
wherein the one or more feature sets include one or more timestamps associated with one or more commit times associated with the impending code change, and
wherein the one or more feature sets include calendar data associated with the impending code change;
provide the one or more feature sets to a machine learning model trained to indicate one or more recommended actions for the impending code change based on a predicted risk associated with deploying the impending code change,
wherein the predicted risk associated with deploying the impending code change is based on one or more of a first probability that deploying the impending code change will cause code breakage, a second probability that deploying the impending code change will cause functionality breakage, or a third probability that deploying the impending code change will cause support issues, and
wherein the predicted risk is based on cross-referencing the one or more commit times with the calendar data;
trigger, based on providing the one or more feature sets to the machine learning model and based on determining application program interface (API) endpoints that will be impacted by the impending code change the one or more recommended actions for the impending code change;
configure, based on triggering the one or more recommended actions, a deployment pattern associated with deploying the impending code change;
deploy, based on the deployment pattern, the impending code change to change the existing code; and
control, based on the deployment pattern, network traffic routed to new instances associated with the impending code change.