Method and system of generating a call agent avatar using artificial intelligence

• US 10,387,963 B1
• Filed: 09/24/2015
• Issued: 08/20/2019
• Est. Priority Date: 09/24/2015
• Status: Active Grant

• Current Assignee: State Farm Mutual Automobile Insurance Company (State Farm)
• Original Assignee: State Farm Mutual Automobile Insurance Company (State Farm)
• Inventors: Leise, William J., Graff, Douglas Albert, Mullen, Christina P.
• Primary Examiner(s): Borlinghaus, Jason
• Assistant Examiner(s): Alladin, Ambreen A.
• Application Number:
  US14/864,645
• Time in Patent Office: 1,426 Days
• Field of Search: 705 4
• CPC Class Codes: G06Q 40/08 Insurance
• CPC Class Codes: H04M 2203/1025 of avatars
• CPC Class Codes: H04M 2203/253 where a visual mode is used instead of a voice mode
• CPC Class Codes: H04M 3/42 Systems providing special services or facilities to subscribers specially adapted for wireless communication networks H04W4/00
• CPC Class Codes: H04M 3/527 Centralised call answering arrangements not requiring operator intervention

Abstract

A method and system may generate a call agent avatar for responding to insurance-related inquiries. A user'"'"'s client device such as a smart-phone, smart watch, or tablet computer, may receive an insurance-related inquiry or banking-related inquiry from the user. Several data sources may be utilized to automatically determine a response to the insurance-related inquiry or banking-related inquiry, and a call agent avatar may be displayed on the user'"'"'s client device while the response is displayed/played by the client device. The call agent avatar may be animated to move in synchronization with the response so that it appears the call agent avatar is providing the response to the user.

Claims

1. A computer-implemented method for generating a call agent avatar, the method executed by one or more processors programmed to perform the method, the method comprising:

receiving, at one or more processors in a server device, input from a client device of a user including an insurance-related inquiry;

identifying, at the one or more processors, a plurality of requested actions for the one or more processors to perform, wherein each requested action is performed in response to an insurance-related inquiry and includes a set of rules for providing a response to the corresponding insurance-related inquiry, and wherein the response includes insurance-related information;

transforming, by the one or more processors, the input from the user to a requested action of the plurality of requested actions by;

comparing, by the one or more processors, the input to a set of grammar rules;

identifying, by the one or more processors, a plurality of requested actions;

analyzing, by the one or more processors via a grammar module, the input to identify a likelihood that each of the plurality of requested actions corresponds to the input;

assigning, by the one or more processors, a probability to each of the plurality of requested actions based on the likelihood that each requested action corresponds to the input;

ranking, by the one or more processors, each of the plurality of requested actions based on their respective probabilities;

identifying, by the one or more processors, the requested action from the plurality of requested actions based on the comparison, including;

identifying, by the one or more processors, a first requested action of the plurality of requested actions having a highest probability of corresponding to the input based on the comparison of the input to the set of grammar rules;

determining, by the one or more processors, that the first requested action has a likelihood of corresponding to the input that is less than a predetermined threshold likelihood;

requesting, by the one or more processors, additional input from the user; and

determining, by the one or more processors, a second requested action as a new requested action based on the input and the additional input;

performing the new requested action, by the one or more processors and via the set of rules which correspond to the new requested action, including analyzing customer data and insurance data corresponding to the user to identify insurance-related information in response to the new requested action;

providing, by the one or more processors to the client device of the user, a graphical representation of a customer service representative for display on the client device; and

providing, by the one or more processors to the client device, a response to the insurance-related inquiry for presentation on the client device, including animating the graphical representation of the customer service representative to move in synchronization with the response, wherein the response includes the insurance-related information.

2. The method of claim 1, wherein the input is an audio input and transforming the input to a requested action includes:

transcribing, by the one or more processors, the audio input into a text input.

3. The method of claim 1, wherein each of the plurality of requested actions is associated with one or more pre-stored responses, the insurance-related information, and the set of rules for determining the insurance-related information in response to the requested action.

4. The method of claim 1, wherein the second requested action is automatically learned as the new requested action based on the input and the additional input such that the additional input is not requested to determine the new requested action when a same input is received at a later time.

5. The method of claim 1, further comprising:

receiving, by the one or more processors, impact characteristics for a vehicle of the user based on sensor data from one or more sensors communicatively coupled to the client device;

identifying, by the one or more processors, that the user has been in a vehicle crash based on the impact characteristics; and

in response to identifying the vehicle crash, causing, by the one or more processors, voice recognition in the client device to be automatically activated to receive the input.

6. The method of claim 1, wherein when the new requested action is to begin a vehicle claims process, the method further comprises:

receiving, by the one or more processors, impact characteristics for a damaged vehicle of the user;

determining, by the one or more processors, a repair complexity level for repairing the damaged vehicle based on the impact characteristics, wherein the repair complexity level represents a degree of difficulty associated with repairing the damaged vehicle; and

selecting, by the one or more processors, a vehicle repair facility for treating the damaged vehicle based on the repair complexity level;

wherein the insurance-related information includes information associated with transporting the damaged vehicle to the selected vehicle repair facility.

7. The method of claim 1, wherein when the new requested action is to provide an amount of a vehicle claim deductible, the method further comprises:

retrieving, by the one or more processors, a customer identifier associated with the user;

retrieving, by the one or more processors, a vehicle loss report associated with the user based on the customer identifier;

retrieving, by the one or more processors, a vehicle insurance policy associated with the user based on the customer identifier; and

determining, by the one or more processors, an amount of the vehicle claim deductible based on the vehicle loss report and the vehicle insurance policy;

wherein the insurance-related information is the amount of the vehicle claim deductible.

8. The method of claim 1, wherein the response is an audio response, and further comprising animating, by the one or more processors, the graphical representation to match the audio response.

9. The method of claim 1, wherein the response is a text response segmented into a plurality of portions, and each portion of the text response is displayed at a different point in time during the animation of the customer service representative.

18. The method of claim 1, wherein animating the graphical representation of the customer service representative to move in synchronization with the response includes animating the graphical representation of the customer service representative to change expression in accordance with a context of the response.

10. A system for generating a call agent avatar, the system comprising:

one or more processors;

a communication network;

a non-transitory computer-readable memory coupled to the one or more processors, and the communication network, and storing thereon instructions that, when executed by the one or more processors, cause the system to;

receive, via the communication network, input from a client device of a user including an insurance-related inquiry;

identify a plurality of requested actions for the one or more processors to perform, wherein each requested action is performed in response to an insurance-related inquiry and includes a set of rules for providing a response to the corresponding insurance-related inquiry, and wherein the response includes insurance-related information;

transform the input from the user to a requested action of the plurality of requested actions by;

comparing the input to a set of grammar rules;

identifying a plurality of requested actions;

analyzing, via a grammar module, the input to identify a likelihood that each of the plurality of requested actions corresponds to the input;

assigning a probability to each of the plurality of requested actions based on the likelihood that each requested action corresponds to the input;

ranking each of the plurality of requested actions based on their respective probabilities;

identifying the requested action from the plurality of requested actions based on the comparison, including;

identifying a first requested action of the plurality of requested actions having a highest probability of corresponding to the input based on the comparison of the input to the set of grammar rules;

determining that the first requested action has a likelihood of corresponding to the input that is less than a predetermined threshold likelihood;

requesting additional input from the user; and

determining a second requested action as a new requested action based on the input and the additional input;

perform the new requested action, via the set of rules which correspond to the new requested action, including analyzing customer data and insurance data corresponding to the user to identify insurance-related information in response to the new requested action;

provide, via the communication network to the client device of the user, a graphical representation of a customer service representative for display on the client device; and

provide, via the communication network to the client device, a response to the insurance-related inquiry for presentation on the client device, including animating the graphical representation of the customer service representative to move in synchronization with the response, wherein the response includes the insurance-related information.

11. The system of claim 10, wherein the input is an audio input and to transform the input to a requested action, the instructions cause the system to:

transcribe the audio input into a text input.

12. The system of claim 10, wherein the second requested action is automatically learned as the new requested action based on the input and the additional input such that the additional input is not requested to determine the new requested action when a same input is received at a later time.

13. The system of claim 10, wherein the instructions further cause the system to:

receive, via the communication network, impact characteristics for a vehicle of the user based on sensor data from one or more sensors communicatively coupled to the client device;

identify that the user has been in a vehicle crash based on the impact characteristics; and

in response to identifying the vehicle crash, cause, via the communication network, voice recognition in the client device to be automatically activated to receive the input.

14. The system of claim 10, wherein when the new requested action is to begin a vehicle claims process, the instructions further cause the system to:

receive, via the communication network, impact characteristics for a damaged vehicle of the user;

determine a repair complexity level for treating the damaged vehicle based on the impact characteristics, wherein the repair complexity level represents a degree of difficulty associated with treating the damaged vehicle; and

select a vehicle repair facility for treating the damaged vehicle based on the repair complexity level;

wherein the insurance-related information includes information associated with transporting the damaged vehicle to the selected vehicle repair facility.

15. The system of claim 10, wherein when the new requested action is to provide an amount of a vehicle claim deductible, the instructions further cause the system to:

retrieve a customer identifier associated with the user;

retrieve a vehicle loss report associated with the user based on the customer identifier;

retrieve a vehicle insurance policy associated with the user based on the customer identifier; and

determine an amount of the vehicle claim deductible based on the vehicle loss report and the vehicle insurance policy;

wherein the insurance-related information is the amount of the vehicle claim deductible.

16. The system of claim 10, wherein the response is an audio response, and further comprising animating, by the one or more processors, the graphical representation to match the audio response.

17. The system of claim 10, wherein the response is a text response segmented into a plurality of portions, and each portion of the text response is displayed at a different point in time during the animation of the customer service representative.