Granted Patent B2

US 12,243,045 · App. 18/318,227 · Granted Mar 4, 2025

Inventors: Elizabeth Qian (Long Island, NY); Robert Fenster (Reston, VA); David Huang (Henrico, VA)

Assignee: CAPITAL ONE SERVICES, LLC

G06Q20/354G06V40/172G10L17/00

Abstract

Claims (70)

1. A card authentication system comprising:

one or more processors; and

memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, are configured to cause the card authentication system to:

receive, via a user device, an audio file comprising a first voice of a customer;

associate the audio file with a first customer account of the customer;

store the audio file;

receive, via the user device, a first credential token and a voice request to activate a card associated with a second customer account, wherein the voice request comprises a first number associated with the card;

generate, using a first machine learning model, a voice fingerprint confidence score based on comparing a second voice in the voice request to the first voice of the stored audio file;

generate, using a second machine learning model, a device authentication score based on (i) comparing an address from the user device to a stored address of a known user device associated with the second customer account and (ii) determining whether the user device and the known user device are connected to a shared communication network;

determine whether the first credential token is valid by comparing the first credential token to a second credential token associated with the user device;

determine whether the voice request is valid based at least in part on the determination of whether the first credential token is valid and by:

determining whether the first number matches a stored second number associated with the card;

responsive to determining the first number matches the stored second number, determining whether the device authentication score exceeds a first predetermined threshold; and

responsive to determining the device authentication score exceeds the first predetermined threshold, determining whether the voice fingerprint confidence score satisfies a second predetermined threshold; and

activate the card based on the determination that the voice request is valid.

2. The system of claim 1 , wherein the instructions are further configured to cause the card authentication system to:

receive a request for the second credential token associated with the user device;

generate the second credential token; and

store the second credential token.

3. The system of claim 1 , wherein determining whether the first number matches the stored second number is conducted using a third machine learning model.

4. The system of claim 1 , wherein the instructions are further configured to cause the card authentication system to:

transmit, via the user device, a second audio file based on the determination that the voice request is not valid.

5. The system of claim 4 , wherein the second audio file comprises a request that the customer submit a new voice request.

6. The system of claim 4 , wherein the second audio file comprises a request that the customer be contacted via a stored phone number to complete activation of the card.

7. The system of claim 1 , wherein generating the device authentication score is further based on determining whether a location of the user device is within a predetermined distance of a known location of a second customer associated with the second customer account.

8. The system of claim 1 , wherein comparing the second voice in the voice request to the first voice of the stored audio file comprises comparing one or more respective voice features, the one or more respective voice features comprising one or more of pitch, tone, cadence, rate, volume, pace, pause, fluency, intensity, articulation, enunciation, or combinations thereof.

9. The system of claim 1 , wherein the instructions are further configured to cause the card authentication system to:

receive image data via the user device;

extract an image of a speaker of the voice request from the image data; and

determine whether the image of the speaker matches a stored image of a user associated with the second customer account.

10. A card authentication system comprising:

one or more processors; and

memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, are configured to cause the card authentication system to:

receive, via a user device, a voice request to activate a card associated with a customer account, wherein the voice request comprises a first number associated with the card;

generate, using a first machine learning model, a voice fingerprint confidence score based on comparing a first voice in the voice request to a second voice in a stored audio file associated with a customer of the customer account;

generate, using a second machine learning model, a device authentication score based on (i) comparing an address from the user device to a stored address of a known user device associated with the customer account and (ii) determining whether the user device and the known user device are connected to a shared communication network;

determine whether the voice request is valid by:

determining whether the first number matches a stored second number associated with the card;

responsive to determining the first number matches the stored second number, determining whether the device authentication score exceeds a first predetermined threshold; and

responsive to determining the device authentication score exceeds the first predetermined threshold, determining whether the voice fingerprint confidence score satisfies a second predetermined threshold; and

activate the card based on the determination that the voice request is valid.

11. The system of claim 10 , wherein the instructions are further configured to cause the card authentication system to:

receive the audio file via the user device;

associate the audio file with the customer account; and

store the audio file.

12. The system of claim 10 , wherein determining whether the first number matches the stored second number is conducted using a third machine learning model.

13. The system of claim 10 , wherein the instructions are further configured to cause the card authentication system to:

transmit, via the user device, a second audio file based on the determination that the voice request is not valid,

wherein the second audio file comprises a request that the customer be contacted via a stored phone number to complete activation of the card.

14. The system of claim 10 , wherein generating the device authentication score is further based on determining whether a location of the user device is within a predetermined distance of a known location of a second customer associated with the customer account.

15. The system of claim 10 , wherein comparing the first voice in the voice request to a second voice in a stored audio file comprises comparing one or more respective voice features, the one or more respective voice features comprising one or more of pitch, tone, cadence, rate, volume, pace, pause, fluency, intensity, articulation, enunciation, or combinations thereof.

16. A card authentication system comprising:

one or more processors; and

memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, are configured to cause the card authentication system to:

receive a request for a first credential token associated with a user device;

generate the first credential token;

store the first credential token;

receive, via the user device, a second credential token and a voice request to activate a card associated with a customer account, wherein the voice request comprises a first number associated with the card;

generate, using a first machine learning model, a voice fingerprint confidence score based on comparing a voice in the voice request to a stored voice associated with the customer account;

determine whether the second credential token is valid by comparing the first credential token to the second credential token;

determine whether the voice request is valid based at least in part on the determination of whether the second credential token is valid and by:

determining whether the user device and a known user device are connected to a shared communication network;

determining whether the first number matches a stored second number associated with the card; and

responsive to determining the first number matches the stored second number, determining whether the voice fingerprint confidence score satisfies a second predetermined threshold; and

activate the card based on the determination that the voice request is valid.

17. The system of claim 16 , wherein the instructions are further configured to cause the card authentication system to:

transmit, via the user device, an audio file based on the determination that the voice request is not valid,

wherein the audio file comprises a request that a customer associated with the customer account be contacted via a stored phone number to complete activation of the card.

18. The system of claim 16 , wherein determining whether the first number matches the stored second number is conducted using a second machine learning model.

19. The system of claim 1 , wherein the user device comprises a smart speaker.

Assignments (1)

Continuity (1)

References Cited (13)