Granted Patent B2

US 12,367,519 · App. 17/530,183 · Granted Jul 22, 2025

Inventors: Michael Mossoba (Great Falls, VA); Abdelkader M'Hamed Benkreira (New York, NY); Joshua Edwards (Philadelphia, PA)

Assignee: Capital One Services, LLC

G06Q30/0631G06Q30/012

Abstract

Claims (66)

1. A method, comprising:

activating an integrated application, wherein the integrated application is integrated with a browser of a user device;

detecting, by the integrated application, a potential purchase of a first product by a user, wherein the potential purchase is conducted via the user device;

retrieving, by the integrated application, warranty claims data of the first product from a warranty database, wherein the integrated application is configured to interface with the warranty database;

determining, by the integrated application, a set of customers which are similar to the user by comparing a similarity score of one or more prior customers against a similarity score of the user;

transmitting, by the integrated application, instructions for modifying a machine-learning model using transaction data of the set of customers which are similar to the user, thereby producing a modified machine-learning model;

causing analysis, by the integrated application, via the modified machine-learning model, of a subset of the warranty claims data to determine a purchase recommendation regarding the first product,

wherein the purchase recommendation includes:

a recommendation to purchase a second product other than the first product, and

a recommendation not to purchase the first product;

blocking, by the integrated application, via the user device and prior to the purchase of the first product, the purchase of the first product based on an output of the modified machine-learning model;

altering a webpage presented on the user device by inactivating a portion of the webpage such that a purchase of the first product cannot be completed; and

causing to output, via a graphical user interface, the altered webpage.

2. The method of claim 1 , the analysis comprising:

generating the recommendation via the modified machine-learning model.

3. The method of claim 2 , further comprising:

wherein the machine-learning model is trained using data associated with a set of customers that exceed a similarity score threshold with the user.

4. The method of claim 3 ,

wherein the second product is determined via the modified machine-learning model with returns data and transaction data of the set of customers.

5. The method of claim 1 , the detecting comprising:

identifying patterns on one or more webpages that can be accessed by the user device; and

detecting a present pattern matches an identified pattern to determine a potential purchase.

6. The method of claim 1 , further comprising:

generating a pop-up webpage window, a browser redirect, or information overlay to present the purchase recommendation.

7. A system, comprising:

a memory storing processor readable instructions; and

at least one processor configured to access the memory and execute the processor readable instructions, which when executed by the at least one processor, configures the at least one processor to perform a plurality of functions, including functions for:

activating an integrated application, wherein the integrated application is integrated within a browser of a user device;

identifying, by the integrated application, a first product that a user may purchase using a user device;

retrieving, by the integrated application, warranty claims data of the first product from a warranty database, wherein the integrated application is configured to interface with the warranty database;

determining, by the integrated application, a set of customers which are similar to the user by comparing a similarity score of one or more prior customers against a similarity score of the user;

transmitting, by the integrated application, instructions for modifying a machine-learning model using transaction data of the set of customers which are similar to the user, thereby producing a modified machine-learning model;

causing analysis, by the integrated application, via the modified machine-learning model, of a subset of the warranty claims data to determine a purchase recommendation regarding the first product, wherein the recommendation includes a recommendation to purchase a second product other than the first product;

blocking, by the integrated application, via the user device and prior to the purchase of the first product, the purchase of the first product based on an output of the modified machine-learning model;

altering a webpage presented on the user device, the altering comprising inactivating a portion of the webpage such that a purchase of the first product cannot be completed; and

causing to output, via a graphical user interface, the altered webpage.

8. The system of claim 7 ,

wherein the integrated application:

monitors the browser for potential purchases; and

interfaces with the warranty database to retrieve the warranty claims data.

9. The system of claim 8 , wherein the functions further include:

identifying patterns on one or more webpages that can be accessed by the user device; and

detecting a present pattern matches an identified pattern via the integrated application to determine a potential purchase.

10. The system of claim 7 , wherein the analysis comprises:

training the machine-learning model via a machine learning technique using the transaction data of the set of customers which are similar to the user; and

generating the recommendation via the modified machine-learning model.

11. The system of claim 10 ,

wherein the machine-learning model has been trained using a set of customers that exceed a similarity score threshold with the user.

12. The system of claim 11 ,

wherein the second product is determined via the modified machine-learning model with returns data and transaction data of the set of customers.

13. The system of claim 7 ,

wherein the purchase recommendation further includes a recommendation to not purchase the first product.

14. A non-transitory computer readable storage medium storing instructions to control one or more processors to perform operations, including:

detecting a first product being considered for purchase via a user device;

activating an integrated application on the device based on the detection;

retrieving, by the integrated application, warranty claims data of the first product being considered for purchase from a warranty database, wherein the integrated application is configured to interface with the warranty database;

determining, by the integrated application, a set of customers which are similar to the user by comparing a similarity score of one or more prior customers against a similarity score of the user;

transmitting instructions, by the integrated application, for modifying a machine-learning model using transaction data of the set of customers which are similar to the user, thereby producing a modified machine-learning model;

causing analysis, by the integrated application, via the modified machine-learning model, of a subset of the warranty claims data to establish a purchase recommendation for purchasing the first product,

wherein the purchase recommendation includes:

a recommendation to purchase a second product other than the first product, and

a recommendation not to purchase the first product;

blocking, by the integrated application, via the user device and prior to the purchase of the first product, the purchase of the first product based on an output of the modified machine-learning model;

altering a webpage presented on the user device, the altering comprising inactivating a portion of the webpage such that a purchase of the first product cannot be completed; and

causing to output, via a graphical user interface, the altered webpage.

15. The computer readable medium of claim 14 , wherein the second product is determined via the modified machine-learning model with returns data and transaction data of a set of customers.

Assignments (1)

Continuity (1)

References Cited (29)