Concurrent Application Testing Platform Using ML-Generated Test Variants
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current application testing techniques are cumbersome, haphazard, and inefficient, leading to inaccurate results and wastage of computing and networking resources due to disjointed testing, making it difficult to identify improvements in customer experience for product and service applications.
Innovation Solution
A machine learning-based testing platform that generates modified applications for concurrent testing, processing parameters and application data to create test applications, assigns user devices to test groups, and provides near-real-time feedback for iterative improvements, conserving resources and enhancing testing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional application testing techniques are used, then testing can be performed, but the process becomes cumbersome and haphazard, leading to inaccurate results and resource wastage
Solution Approach 1:
The patent segments the monolithic application into multiple modular versions that can be tested independently and concurrently. Each modified application version represents a discrete segment of the testing process, allowing systematic comparison without the complexity of testing the entire application at once. This segmentation enables precise measurement of specific modifications while reducing overall testing process complexity.
Solution Approach 2:
The patent creates multiple copies of the application with systematic modifications rather than manually recreating test scenarios. These copied and modified application versions are generated automatically, enabling concurrent testing of multiple hypotheses simultaneously. This approach improves testing accuracy through replication while reducing the cumbersome nature of manual testing processes.
2Productivity
If extensive testing and recoding of web applications is performed to find optimum product presentation, then better customer experience can be achieved, but computing and networking resources are wasted due to disjointed testing
Solution Approach 1:
The patent implements continuous concurrent testing of multiple application modifications simultaneously rather than sequential disjointed testing. Multiple modified applications are tested in parallel, maintaining continuous useful action throughout the testing process. This eliminates resource wastage associated with stopping and starting different test scenarios, while accelerating the rate at which customer experience improvements are identified and implemented.
Solution Approach 2:
The patent systematically changes application parameters (such as product presentation features, UI elements, or functional characteristics) across multiple modified versions to test their impact on customer experience. By varying specific parameters across concurrent test groups rather than performing exhaustive recoding and testing, the system efficiently identifies optimal configurations while minimizing computing resource consumption.
3Measurement precision
If multiple application modifications are tested sequentially, then each modification can be evaluated, but the process is slow and inefficient
Solution Approach 1:
The patent merges multiple sequential testing operations into a single concurrent testing framework. Multiple modified applications are tested simultaneously in the same environment with shared resources, combining what would have been separate sequential testing campaigns. This maintains measurement precision by controlling for environmental variables while dramatically reducing the total time required to evaluate multiple modifications.
Data Source
AI summary
A device may receive parameters to test modifications to an application associated with a product and/or a service, and may process data identifying the parameters and the application, with a machine learning model, to generate test applications for testing corresponding modifications to the application. The device may define test group sizes of test groups for testing the test applications, and may receive, from user devices, requests for accessing the application. The device may assign, based on the test group sizes, sets of the user devices to the test groups for testing the test applications, and may provide the test applications concurrently to the corresponding sets of the user devices based on the test groups. The device may receive, from the corresponding sets of the user devices, feedback associated with the test applications, and may perform one or more actions based on the feedback.