US 12,169,804 B1
System and method for dynamic context sensitive guidance
Eric Jason Grenet, Boerne, TX (US); Brian Christopher Hawes, San Antonio, TX (US); Paul Christopher Blanchard, San Antonio, TX (US); Phillip E. Marks, San Antonio, TX (US); Jeffrey Walton Easley, San Antonio, TX (US); Douglas Austin Johnson, San Antonio, TX (US); Katrina Marie Zell, San Antonio, TX (US); Julia Yilan Kennedy, Austin, TX (US); Lawrence Paul McDermott, Jr., San Antonio, TX (US); David Morley, San Antonio, TX (US); and Christopher Collin Campbell, San Antonio, TX (US)
Assigned to United Services Automobile Association (USAA), San Antonio, TX (US)
Filed by UIPCO, LLC, San Antonio, TX (US)
Filed on Mar. 31, 2022, as Appl. No. 17/657,407.
Claims priority of provisional application 63/169,076, filed on Mar. 31, 2021.
Int. Cl. G06Q 10/0639 (2023.01); G06F 9/451 (2018.01); G06Q 10/0631 (2023.01)
CPC G06Q 10/0639 (2013.01) [G06F 9/453 (2018.02); G06Q 10/063112 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method of dynamically modifying information presented to a user of an application, the method comprising:
generating, at a first time and by a dynamic response system associated with a first application running on a first computing device, first content for a first message related to a first in-app task to be performed via the first application;
receiving, at a second time from the first application and at a task performance assessor of the dynamic response system, first activity data for a first user corresponding to a successful completion of the first in-app task;
generating, at the task performance assessor, a first efficiency rating describing a performance of the first user in completing the first in-app task based on the first activity data;
determining, at the dynamic response system, that the first efficiency rating is above a first threshold and responsively inputting the first activity data as training data to a task-specific deep learning network model;
identifying, via the task-specific deep learning network model, patterns in the first activity data that increased a likelihood of an efficient performance of the first in-app task:
generating, at a third time and by the task-specific deep learning network model, second content to replace the first content based at least on the identified patterns;
receiving, at a fourth time and at a second application running on a second computing device, second activity data for a second user corresponding to a launch of the first in-app task to be performed via the second application; and
automatically presenting, via the second application, the first message including the second content to the second user.