US 12,271,279 B2
System and method for autonomous testing, machine-learning model-supervised prioritization, and randomized workflow generation
Ranjeet Joseph Kumar Anthonappa, Bengaluru (IN); Venkata Nageswara Rao Desaraju, Bangalore (IN); Sneha Raveendran, Bengaluru (IN); and Sudarshan Babu Kotapati, Bengaluru (IN)
Assigned to Cerner Innovation, Inc., Kansas City, MO (US)
Filed by Cerner Innovation, Inc., Kansas City, MO (US)
Filed on Apr. 16, 2024, as Appl. No. 18/637,270.
Application 18/637,270 is a continuation of application No. 17/487,484, filed on Sep. 28, 2021, granted, now 11,994,966.
Prior Publication US 2024/0264913 A1, Aug. 8, 2024
Int. Cl. G06F 11/22 (2006.01); G06F 11/263 (2006.01); G06N 20/00 (2019.01)
CPC G06F 11/2273 (2013.01) [G06F 11/263 (2013.01); G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
training a machine learning prediction model to identify patterns associated with a type of problem occurring for at least one of a step in a workflow or a specific order of steps in a workflow, based on at least one training dataset comprising a plurality of historical workflows and code defect tags associated with the plurality of historical workflows;
receiving a primary workflow having a plurality of steps organized in a sequence;
generating a set of test workflows having randomized the sequence of at least one step in the plurality of steps of the primary workflow;
identifying, by the machine learning prediction model, a subset of test workflows, within the set of test workflows, that meet or exceed a matching threshold to a pattern associated with a type of problem in a historical workflow;
communicating the subset of test workflows for display in a graphical user interface; and
retraining the machine learning prediction model at least by iteratively applying the machine learning prediction model to additional training datasets and updating the machine learning prediction model based on results generated by iteratively applying the machine learning prediction model to the additional training datasets;
wherein the method is performed by at least one device including a hardware processor.