Granted Patent B1

US 12,744,848 · App. 18/954,297 · Granted Sep 22, 2026

Inventors: Oscar Roberto Tijerina (San Antonio, TX); Edward Manuel Alonzo, III (San Antonio, TX); Daniel Nathan Kalinoski (San Antonio, TX); Keegan Patrick Hayes (Whitestown, IN); Octavia Lloyd (Riverview, FL); Oscar Guerra (San Antonio, TX)

Assignee: United Services Automobile Association (USAA)

H04M3/5191G10L15/183G10L15/22G10L15/30H04M3/5175H04L51/02H04M2201/39H04M2201/40

Abstract

Claims (88)

1 . A method for configuring smart digital agent(s) for interactions over a user interaction lifecycle, the method comprising:

receiving, via interactions with a live agent, selections of one or more smart digital agents from among a plurality of smart digital agents for a user interaction, wherein the plurality of smart digital agents are categorized based on user issue types;

providing, based on input from the live agent, configuration data for the user interaction, wherein the configuration data comprises live agent controls with respect to one or more user issue types the selected smart digital agent is configured to resolve, the live agent controls comprising issue facts or smart digital agent instruction; and

triggering a dialogue between the selected smart digital agent and the user, wherein,

the selected smart digital agent navigates a predefined issue workflow by analyzing audio input from the user and generating speech responses to the user, and

the selected smart digital agent:

omits one or more portions of the predefined issue workflow based on the issue facts and/or the smart digital agent instructions provided to the smart digital agent via the live agent controls, and/or

augments the predefined issue workflow with one or more additional portions based on the issue facts and/or the smart digital agent instructions provided to the smart digital agent via the live agent controls.

2 . The method of claim 1 , wherein the input, provided by the live agent used to generate the live agent controls, is provided by the live agent based at least in part on a live discussion with the user that is conducted prior to the triggered dialogue.

3 . The method of claim 1 , wherein,

the predefined issue workflow comprises user or issue variables that the smart digital agent is configured to resolve via the dialogue;

the live agent controls comprise at least one issue fact that resolves at least one user or issue variable, and

the smart digital agent is configured to:

omit a portion of the predefined issue workflow that corresponds to the at least one issue variable.

4 . The method of claim 1 , wherein,

the predefined issue workflow comprises user or issue variables that the smart digital agent is configured to resolve via the dialogue;

the live agent controls comprise at least smart digital agent instruction that introduces a new user or issue variable, and

the smart digital agent is configured to:

augment a portion of the predefined issue workflow to resolve the introduced new user or issue variable.

5 . The method of claim 1 , wherein the smart digital agent is configured to:

ingest the live agent controls;

compare the ingested live agent controls to the predefined issue workflow;

identify one or more portions of the predefined issue workflow based on the comparison; and

perform one or more of:

omitting the identified one or more portions of the predefined issue workflow; or

augmenting the identified one or more portions of the predefined issue workflow with one or more additional workflow portions.

6 . The method of claim 5 , wherein,

the predefined workflow comprises a workflow structure with workflow states and workflow transitions, and

navigating the predefined issue workflow by the smart digital agent comprises generating the speech responses by progressing through the workflow states according to the workflow transitions.

7 . The method of claim 6 , wherein

the one or more portions of the predefined issue workflow identified based on the comparison comprise one or more states and one or more state transitions that match the live agent controls, and

the smart digital agent is configured to perform one or more of:

omitting the one or more states and one or more state transitions that match the live agent controls; or

augmenting the one or more states and one or more state transitions that match the live agent controls with one or more additional workflow portions.

8 . The method of claim 5 , wherein,

the predefined workflow comprises a series of organized language prompts, and

navigating the predefined issue workflow by the smart digital agent comprises providing the organized language prompts to a generative machine learning model to generate the speech responses.

9 . The method of claim 8 , wherein

the one or more portions of the predefined issue workflow identified based on the comparison comprise one or more language prompts that match the live agent controls, and

the smart digital agent is configured to perform one or more of:

omitting the or more language prompts that match the live agent controls; or

augmenting the or more language prompts that match the live agent controls with additional language prompts.

10 . The method of claim 1 , further comprising:

receiving, during the triggered dialogue from a monitoring live agent, one or more real-time live agent controls; and

generating, by the smart digital agent during the triggered dialogue, at least one speech response based on the real-time live agent controls.

11 . The method of claim 1 , further comprising:

receiving, during the triggered dialogue from a monitoring live agent, real-time live agent input that simulates audio input from the user; and

generating, by the smart digital agent during the triggered dialogue, at least one speech response based on the simulated audio input from the user.

12 . The method of claim 11 , wherein the user is provided the real-time live agent input that simulates the audio input from the user, or the user is restricted from the real-time live agent input that simulates the audio input from the user.

13 . The method of claim 1 , further comprising:

calculating, during the triggered dialogue, a user satisfaction score based on a progress of navigating the predefined issue workflow and/or the user audio input; and

prompting, when the user satisfaction score is below a threshold, a live agent monitoring the triggered dialogue to interrupt, wherein the smart digital agent is configured to pause or terminate navigating the predefined issue workflow when the live agent interrupts the dialogue.

14 . The method of claim 1 , further comprising:

after the triggered dialogue, triggering a follow-up dialogue between the user and another smart digital agent configured to navigate a predefined follow-up workflow, wherein the another smart digital agent is selected based on one or more follow-up actions associated with the triggered dialogue or the predefined issue workflow.

15 . A non-transitory computer-readable storage medium storing instructions that, when executed by a computing system, cause the computing system to perform a process for configuring smart digital agent(s) for interactions over a user interaction lifecycle, the process comprising:

receiving, via interactions with a live agent, selections of one or more smart digital agents from among a plurality of smart digital agents for a user interaction, wherein the plurality of smart digital agents are categorized based on user issue types;

providing, based on input from the live agent, configuration data for the user interaction, wherein the configuration data comprises live agent controls with respect to one or more user issue types the selected smart digital agent is configured to resolve, the live agent controls comprising issue facts or smart digital agent instruction; and

triggering a dialogue between the selected smart digital agent and the user, wherein,

the selected smart digital agent navigates a predefined issue workflow by analyzing audio input from the user and generating speech responses to the user, and

the selected smart digital agent:

omits one or more portions of the predefined issue workflow based on the issue facts and/or the smart digital agent instructions provided to the smart digital agent via the live agent controls, and/or

augments the predefined issue workflow with one or more additional portions based on the issue facts and/or the smart digital agent instructions provided to the smart digital agent via the live agent controls.

16 . The non-transitory computer-readable storage medium of claim 15 , wherein,

the predefined issue workflow comprises user or issue variables that the smart digital agent is configured to resolve via the dialogue;

the live agent controls comprise at least one issue fact that resolves at least one user or issue variable, and

the smart digital agent is configured to:

omit a portion of the predefined issue workflow that corresponds to the at least one issue variable.

17 . The non-transitory computer-readable storage medium of claim 15 , wherein,

the predefined issue workflow comprises user or issue variables that the smart digital agent is configured to resolve via the dialogue;

the live agent controls comprise at least smart digital agent instruction that introduces a new user or issue variable, and

the smart digital agent is configured to:

augment a portion of the predefined issue workflow to resolve the introduced new user or issue variable.

18 . The non-transitory computer-readable storage medium of claim 15 , wherein the process further comprises:

receiving, during the triggered dialogue from a monitoring live agent, one or more real-time live agent controls; and

generating, by the smart digital agent during the triggered dialogue, at least one speech response based on the real-time live agent controls.

19 . The non-transitory computer-readable storage medium of claim 15 , wherein the process further comprises:

receiving, during the triggered dialogue from a monitoring live agent, real-time live agent input that simulates audio input from the user; and

generating, by the smart digital agent during the triggered dialogue, at least one speech response based on the simulated audio input from the user.

20 . A computing system for configuring smart digital agent(s) for interactions over a user interaction lifecycle, the computing system comprising:

one or more processors; and

one or more memories storing instructions that, when executed by the one or more processors, cause the computing system to perform a process comprising:

receiving, via interactions with a live agent, selections of one or more smart digital agents from among a plurality of smart digital agents for a user interaction, wherein the plurality of smart digital agents are categorized based on user issue types;

providing, based on input from the live agent, configuration data for the user interaction, wherein the configuration data comprises live agent controls with respect to one or more user issue types the selected smart digital agent is configured to resolve, the live agent controls comprising issue facts or smart digital agent instruction; and

triggering a dialogue between the selected smart digital agent and the user, wherein,

the selected smart digital agent navigates a predefined issue workflow by analyzing audio input from the user and generating speech responses to the user, and

the selected smart digital agent:

omits one or more portions of the predefined issue workflow based on the issue facts and/or the smart digital agent instructions provided to the smart digital agent via the live agent controls, and/or

augments the predefined issue workflow with one or more additional portions based on the issue facts and/or the smart digital agent instructions provided to the smart digital agent via the live agent controls.

Continuity (1)

References Cited (9)