Granted Patent B2

US 12,541,652 · App. 18/598,092 · Granted Feb 3, 2026

Inventors: Manu Kurian (Dallas, TX); Vinesh Patel (London, GB)

Assignee: Bank of America Corporation

G06F40/35G06F21/31G06F40/279G06N3/006H04L51/02

Abstract

Claims (81)

1 . A method for generating an artificial intelligence (“AI”) chatbot having an AI persona customized for a user of an entity, the method comprising:

generating the AI persona for the AI chatbot, the AI persona having customizable user-interaction characteristics that include:

an AI persona tonality;

an AI persona response time; and

an AI persona formality level;

receiving a request from the user to interact with the AI chatbot;

in response to receipt of the request, modifying the AI persona of the AI chatbot from a default state to an initial customized state by selecting, from a plurality of characteristic types, a characteristic type for each of the AI persona tonality, the AI persona response time, and the AI persona formality level, wherein the selecting comprises:

analyzing user data, the user data including a user age, a preferred language of the user, and a user profile; and

selecting an initial characteristic type for each of the customizable user-interaction characteristics based on the analyzing of the user data;

initiating an interaction between the user and the AI chatbot, the AI persona of the AI chatbot being in the initial customized state; and

during a duration of the interaction, in real-time, performing, in parallel, a first monitoring routine, a second monitoring routine and a third monitoring routine, each of the first monitoring routine, the second monitoring routine and the third monitoring routine for identifying an opportunity to modify the initial customized state of the AI persona to a revised customized state, wherein:

the first monitoring routine comprises:

monitoring a tone of user input for identifying a tone that is outside a range of predicted tones; and

in response to a determination that the tone of the user input is outside the range of predicted tones, adjusting the initial customized state of the AI tonality to a revised customized state of the AI tonality;

the second monitoring routine comprises:

monitoring a speed of user input for detecting a speed that is outside a range of predicted speeds; and

in response to a determination that the speed of the user input is outside the range of predicted speeds, adjusting the initial customized state of the AI response time to a revised customized state of the AI response time; and

the third monitoring routine comprises:

monitoring text included in the user input for text including one or more predetermined keywords; and

in response to an identifying of one or more predetermined keywords, adjusting the initial customized state of the AI formality level to a revised customized state of the AI formality level.

2 . The method of claim 1 further comprising storing the characteristic type for each of the AI persona tonality, the AI persona response time, and the AI persona formality level associated with the AI persona at a time of termination of the interaction.

3 . The method of claim 2 further comprising associating a positive tag or a negative tag with each of the characteristic types.

4 . The method of claim 3 wherein the request is a first request, and the method further comprises receiving a second request from the user.

5 . The method of claim 4 further comprising, in response to receiving the second request, modifying the AI persona of the AI chatbot prior to responding to the second request by selecting the characteristic types stored at the time of the termination.

6 . The method of claim 1 wherein the interaction is via a texting chatbot.

7 . The method of claim 6 further comprising, during the interaction, displaying a pop-up notification comprising selectable options for rating each of the customizable user-interaction characteristics of the AI persona.

8 . The method of claim 7 further comprising, in response to a receipt of a negative rating, modifying the initial customized state of the AI persona to the revised customized state.

9 . The method of claim 1 wherein the interaction is via an interactive voice response (“IVR”) system.

10 . The method of claim 1 wherein the interaction between the user and the AI chatbot is displayed on a user interface on a user device.

11 . A method for generating an artificial intelligence (“AI”) chatbot having a persona customized for a user of an entity, the method comprising:

generating an AI persona for the AI chatbot, the AI persona comprising customizable user-interaction characteristics that include:

a AI persona tonality;

a AI persona response time; and

a AI persona formality level;

receiving a first request from the user to interact with the AI chatbot;

in response to receipt of the first request, modifying the AI persona of the AI chatbot from a default state to an initial customized state by selecting, from a plurality of characteristic types, a characteristic type for each of the AI persona tonality, AI persona response time and AI persona formality level, wherein the selecting comprises:

analyzing user data, the user data comprising: a user age, a preferred language of the user, and a user profile; and

selecting an initial characteristic type for each of the customizable user-interaction characteristics based on the analyzing of the user data;

initiating an interaction between the user and the AI persona, the AI persona being in the initial customized state;

during a duration of the interaction, performing, in parallel, a first monitoring routine, a second monitoring routine and a third monitoring routine, each of the first monitoring routine, the second monitoring routine and the third monitoring routine, for identifying an opportunity to modify the initial customized state of the AI persona to a revised customized state, wherein:

the first monitoring routine comprises:

monitoring a tone of user input for identifying a tone that is outside a range of predicted tones; and

in response to a determination that the tone of the user input is outside the range of predicted tones, adjusting the initial customized state of the AI tonality to a revised customized state of the AI tonality;

the second monitoring routine comprises:

monitoring a speed of user input for detecting a speed that is outside a range of predicted speeds; and

in response to a determination that the speed of the user input is outside the range of predicted speeds, adjusting the initial customized state of the AI response time to a revised customized state of the AI response time; and

the third monitoring routine comprises:

monitoring text included in the user input for input of a one or more predetermined keywords; and

in response to an identifying of the one or more predetermined keywords, adjusting the initial customized state of the AI formality level to a revised customized state of the AI formality level;

at a time of termination of the interaction, storing the characteristic type set for each of the customizable user-interaction characteristics at the time of the termination;

receiving a second request from the user to interact with the AI chatbot; and

automatically modifying the default state of the AI persona to the revised customized state at the time of the termination.

12 . The method of claim 11 further comprising, during the interaction following a receipt of the second request, continuously performing the first monitoring routine, the second monitoring routine and the third monitoring routine.

13 . The method of claim 11 wherein the interaction is via a texting chatbot.

14 . The method of claim 11 wherein the interaction is via an interactive voice response (“IVR”) system.

15 . An artificial intelligence (“AI”) chatbot having a dynamic AI persona for interacting with a user of an entity, the AI persona customizable for the user, the AI chatbot comprising:

the AI persona comprising a plurality of customizable user-interaction characteristics, the plurality of customizable user-interaction characteristics comprising:

an AI persona tonality;

an AI persona response time; and

an AI persona formality level;

a receiver configured to receive a request from a user device associated with the user to interact with the AI chatbot;

a processor configured to execute the AI chatbot to include the AI persona for responding to the request by modifying a default AI persona to an initial customized state by selecting, from a plurality of characteristic types, a characteristic type for each of the AI persona tonality, the AI persona response time, and the AI persona formality level, the selecting comprising:

using a data analytics module to analyze user data from a database, the database stored at an entity server, the processor in electronic communication with the entity server, the user data comprising: a user age, a preferred language of the user, and a user profile; and

based on the analyzing, select an initial characteristic type for each of the customizable user-interaction characteristics; and

the processor further configured to:

initiate an interaction between the user and the AI chatbot wherein the AI persona of the AI chatbot being in the initial customized state; and

during a duration of the interaction, perform, in parallel a first monitoring routine, a second monitoring routine and a third monitoring routine, each of the first monitoring routine, the second monitoring routine and the third monitoring routine, for identifying an opportunity to modify the initial customized state of the AI persona to a revised customized state, wherein:

the first monitoring routine comprises:

monitoring a tone of user input for identifying a tone that is outside a range of predicted tones; and

in response to a determination that the tone of the user input is outside the range of predicted tones, adjusting the initial customized state of the AI tonality to a revised customized state of the AI tonality;

the second monitoring routine comprises:

monitoring a speed of user input for detecting a speed that is outside a range of predicted speeds; and

in response to a determination that the speed of the user input is outside the range of predicted speeds, adjusting the initial customized state of the AI response time to a revised customized state of the AI response time; and

the third monitoring routine comprises:

monitoring text included in the user input for input of a one or more predetermined keywords; and

in response to an identifying of the one or more predetermined keywords, adjusting the initial customized state of the AI formality level to a revised customized state of the AI formality level.

16 . The AI chatbot of claim 15 wherein the processor is further configured to store the characteristic type for each of the AI persona tonality, the AI persona response time, and the AI persona formality level associated with the AI persona at a time of termination of the interaction.

17 . The AI chatbot of claim 16 wherein the processor is further configured to associate a positive tag or a negative tag with each of the characteristic types.

18 . The AI chatbot of claim 16 further comprising an authentication engine configured to evaluate an authentication request received via a user interface and determine whether to authenticate the user or deny authentication.

19 . The AI chatbot of claim 16 wherein the user device is a smartphone.

20 . The AI chatbot of claim 19 wherein the interaction on the user device is one of a voice chat or text chat.

Assignments (1)

Continuity (1)

References Cited (19)