Granted Patent B1

US 12,730,620 · App. 19/411,760 · Granted Sep 8, 2026

Inventor: Dharmayu Purohit (Dublin, IE)

Assignee: Wesco Digital Solutions (Ireland) Limited

G06F8/60G06F8/71G06F11/3668

Abstract

Claims (61)

1 . A computer-implemented method for dynamically orchestrating a deployment pipeline for deploying a cloud application, wherein the deployment pipeline comprises deployment stages each configured to perform one or more deployment operations, the computer-implemented method comprising:

monitoring operational metrics of the deployment pipeline, the operational metrics indicative of an operational behavior of the deployment pipeline;

determining deployment entropy values indicative of a disorder level within the deployment pipeline, wherein the deployment entropy values are based on temporal variances in the monitored operational metrics associated with the deployment stages of the deployment pipeline;

evaluating the deployment entropy values against entropy bands associated with a set of action-selection criteria; and

executing an adaptive pipeline pathing of the cloud application through the plurality of deployment stages by selecting and executing deployment control actions based on the evaluation of the deployment entropy values against the entropy bands;

wherein computing the deployment entropy values comprises applying an equation defined as:

DEV

=

(

CD

+

EI

+

DR

+

CO

)

/

N

wherein DEV is the Deployment Entropy value, CD is Configuration Drift Value, EI is an environmental inconsistency value, DR is a deployment randomness value, CO is a configuration obsolescence value, and N is a normalizing constant greater than zero.

2 . The computer-implemented method of claim 1 , wherein the operational metrics comprise at least one of:

a deployment latency value for each deployment stage; or

a failure rate computed based on success or error outcomes associated with the deployments stages of the deployment pipeline.

3 . The computer-implemented method of claim 1 , wherein the deployment control actions comprise at least one of:

adjusting a concurrency level for deployment tasks;

modifying a retry interval for failure-prone stages; or

reordering execution of non-dependent deployment stages to reduce end-to-end latency.

4 . The computer-implemented method of claim 1 , wherein determining the deployment entropy values comprises:

determining a configuration drift value indicative of inconsistencies across multiple of the deployment stages of the deployment pipeline.

5 . The computer-implemented method of claim 1 , wherein determining the deployment entropy values comprises:

determining an environmental inconsistency value (EI) indicative of defects or test failures repeating in one or more of the deployment stages of the deployment pipeline.

6 . The computer-implemented method of claim 1 , wherein determining the deployment entropy values comprises:

determining a deployment randomness value (DR) indicative of a count of deployments that fail reproducibility from a common source commit.

7 . The computer-implemented method of claim 1 , wherein determining the deployment entropy values comprises:

determining a configuration obsolescence value (CO) indicative of outdated, hardcoded, or deprecated configuration patterns in the deployment pipeline.

8 . The computer-implemented method of claim 1 , wherein the deployment control actions comprise at least one of:

pausing deployment;

executing defined and quantified retries;

accelerating deployment;

initiating rollback to a prior stable deployment stage;

adjusting deployment parameters for subsequent deployment stages; or

invoking a canary deployment for a subset of target environments based on a deployment entropy value deviating from a dynamically adjustable threshold by a first margin.

9 . The computer-implemented method of claim 1 , wherein the monitored operational metrics are indicative of at least one of:

inconsistencies across multiple of the deployment stages of the deployment pipeline;

defects or test failures repeating in one or more of the deployment stages of the deployment pipeline;

a count of deployments that fail reproducibility from a common source commit; or

outdated, hardcoded, or deprecated configuration patterns in the deployment pipeline.

10 . The computer-implemented method of claim 1 , further comprising:

obtaining dependency data indicative of relationships among components of the deployment pipeline; and

determining a coupling value based on the dependency data; and

wherein executing an adaptive pipeline pathing of the cloud application is further based on the coupling value.

11 . The computer-implemented method of claim 10 , wherein the dependency data comprises:

a direct dependency parameter indicative of direct service-to-service or service-to-tool references among components of the deployment pipeline, wherein the direct dependency parameter is based on the obtained dependency data.

12 . The computer-implemented method of claim 10 , wherein the dependency data comprises:

an environment touch parameter indicative of distinct deployment environments affected by changes to at least one of the components of the deployment pipeline.

13 . The computer-implemented method of claim 10 , wherein the dependency data comprises:

a test linkage parameter indicative of integration or end-to-end test cases referencing two or more of the components of the deployment pipeline.

14 . The computer-implemented method of claim 10 , wherein the dependency data comprises:

an infrastructure component parameter indicative of infrastructure resources, including at least one of databases, message buses, or storage buckets, associated with the components of the deployment pipeline.

15 . The computer-implemented method of claim 10 , wherein the dependency data comprises:

a configuration alignment parameter indicative of shared configuration parameters or feature-flag values concurrently referenced by two or more of the components of the deployment pipeline.

Assignments (1)

References Cited (9)