US 11,855,770 B2
Authentication control based on previous actions
Prateek Gupta, Bellerose, NY (US); Samuel Wu, Brooklyn, NY (US); Zachary Wyman, New York, NY (US); and Ramiro Ordonez, Ho Ho Kus, NJ (US)
Assigned to Capital One Services, LLC, McLean, VA (US)
Filed by Capital One Services, LLC, McLean, VA (US)
Filed on Nov. 30, 2021, as Appl. No. 17/538,769.
Application 17/538,769 is a continuation of application No. 17/538,327, filed on Nov. 30, 2021.
Application 17/538,327 is a continuation of application No. 17/533,618, filed on Nov. 23, 2021.
Application 17/533,618 is a continuation of application No. 17/533,773, filed on Nov. 23, 2021.
Prior Publication US 2023/0164075 A1, May 25, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. H04L 67/5682 (2022.01); G06N 20/00 (2019.01); H04L 45/74 (2022.01); H04L 47/28 (2022.01); H04L 49/20 (2022.01); H04M 3/51 (2006.01)
CPC H04L 67/5682 (2022.05) [G06N 20/00 (2019.01); H04L 45/742 (2013.01); H04L 47/286 (2013.01); H04L 49/206 (2013.01); H04M 3/5183 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
receiving, by a server and from an application of a remote user device, user interaction information;
predicting, using a machine-learning model trained on previous interactions of other users with their applications on their user devices and with call centers and based on the received user interaction information, a likelihood of a user of the application to contact a call center with an inquiry from the user;
populating a call center cache based on a prediction that the user will contact the call center;
authenticating the user based on the prediction for the user;
receiving a result of a user interaction between the user and the call center;
storing, in an application cache of a server and with a time-to-live (TTL) value, first information for the application;
monitoring a data stream of incoming information;
detecting in the data stream, second information associated with user;
storing, based on the detecting the second information, the second information in the application cache of the server, wherein the second information overwrites the first information and refreshes the TTL value;
retraining the machine-learning model, wherein the retraining of the machine-learning model is based on:
the previous interactions of the other users with their applications on their user devices and with the call centers;
the received user interaction information; and
the result of the user interaction between the user and the call center; and
repopulating, based on the retraining of the machine-learning model, an updated prediction that the user will contact the call center, and the detection of the second information in the data stream, the call center cache.