# US 2025/0094723 A1 · Negotiation training framework using generative artificial intelligence

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0094723 A1
- Document ID: US20250094723A1
- Status: Published application
- Kind: A1
- Application number: 18/490,921
- Filed: 2023-10-20
- Published: 2025-03-20
- Earliest claimed: 2023-09-20
- Assignees: Insight Direct USA, Inc.
- Applicants: Insight Direct USA, Inc.
- Inventors: Amol Ajgaonkar; Joseph Raymond Flynn
- Web page: https://patentgrep.com/patent/US20250094723A1

## CPC

- G06F 40/35: Discourse or dialogue representation
- G06F 40/30: Semantic analysis
- G06Q 10/06398: {Performance of employee with respect to a job function}
- G06T 13/40: of characters, e.g. humans, animals or virtual beings
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A system for training a user for a conversational encounter with a customer receives encounter-defining parameters via a user interface. The encounter-defining parameters are descriptive of attributes of the conversational encounter. The system formats the encounter-defining parameter for transmission to a large language model (LLM). The LLM receives the instructions and outputs a conversational opening viewable by a user via a user interface. The user then responds to the conversational opening and iteratively converses with the LLM. After a defined maximum number of iterations have been reached, the LLM provides an evaluation to the user via the user interface. The evaluation is indicative of the evaluated outcome of the conversational encounter, including at least one of a score, a summary of the encounter, a likelihood that a deal is reached, and suggested improvements.

## Claims

### Claim 1

A system for using generative artificial intelligence (AI) for training a user for a conversational encounter with a customer, the system comprising:
a processor;
an input/output device operably connected to the processor, the input/output device configured to receive user inputs and to display outputs at a user interface; and
computer-readable memory operably connected to the processor, the computer-readable memory encoded with instructions that, when executed by the processor, cause the system to:
receive, from the user interface, a set of encounter-defining parameters, the encounter-defining parameters being descriptive of attributes of the conversational encounter;
format the set of encounter-defining parameters into a first large language model input, the first large language model input being suitable for processing by a large language model;
provide the first large language model input to the large language model;
receive a first large language model output from the large language model based upon the first large language model input;
provide the first large language model output to the user via the user interface;
iteratively converse with the user, via the user interface, for a defined number of iterations, wherein iteratively conversing with the user includes the steps of:
receiving a user response via the user interface;
providing the user response to the large language model;
receiving an additional large language model output from the large language model; and
providing the additional large language model output to the user via the user interface;
provide an evaluation command to the large language model after the defined number of iterations is reached;
receive a final evaluation from the large language model, the final evaluation being indicative of an evaluated outcome of the conversational encounter; and
display the final evaluation to the user via the user interface.

### Claim 2 (depends on 1)

The system of claim 1, wherein the first large language model output provides a conversational opening statement from the customer that adheres to the set of encounter-defining parameters.

### Claim 3 (depends on 1)

The system of claim 1, wherein the set of encounter-defining parameters include a set of customer-defining parameters and a set of situation-defining parameters.

### Claim 4 (depends on 3)

The system of claim 3, wherein the set of customer-defining parameters include at least one of a personality, a mood, an industry, and a customer role, a team requirement, and a business requirement.

### Claim 5 (depends on 3)

The system of claim 3, wherein the set of situation-defining parameters include at least one of a team requirement and a business requirement.

### Claim 6 (depends on 1)

The system of claim 1, wherein:
the first large language model input includes instructions to evaluate a direction of the user response that is indicative of the relevance of the user response; and
the direction of the user response is displayed to the user via the user interface.

### Claim 7 (depends on 1)

The system of claim 1, wherein the final evaluation includes a numerical score on a predefined scale.

### Claim 8 (depends on 1)

The system of claim 1, wherein the final evaluation includes a summary of the user response.

### Claim 9 (depends on 1)

The system of claim 1, wherein the final evaluation includes a deal result based upon the user response, wherein the deal result is indicative of whether a deal is accepted or rejected.

### Claim 10 (depends on 1)

The system of claim 1, wherein the final evaluation includes one or more suggested improvements to the user response.

### Claim 11 (depends on 1)

The system of claim 1, wherein iteratively conversing with the user further includes:
transmitting a conversation history to the large language model, wherein the conversation history includes prior iterations of the user response and prior iterations of the additional large language model output.

### Claim 12 (depends on 1)

The system of claim 1, wherein iteratively conversing with the user further includes:
transmitting the user response to a sentiment analysis module, wherein the sentiment analysis module produces an emotional index of the user response.

### Claim 13 (depends on 12)

The system of claim 12, wherein iteratively conversing with the user further includes:
transmitting the emotional index of the user response from the sentiment analysis module to the large language model.

### Claim 14 (depends on 13)

The system of claim 13, wherein the first large language model input includes an instruction to respond to the user input based upon the emotional index of the user response.

### Claim 15 (depends on 14)

The system of claim 14, wherein the final evaluation includes a summary of the emotional index of the user response.

### Claim 16 (depends on 15)

The system of claim 15, wherein the computer-readable memory is further encoded with instructions that, when executed by the one or more processors, cause the system to:
generate a virtual reality avatar, wherein the virtual reality avatar is configured to emulate the customer as defined by the encounter-defining parameters; and
display the virtual reality avatar to the user.

### Claim 17 (depends on 16)

The system of claim 16, wherein the virtual reality avatar is configured to display a corresponding response emotion to the user based upon the emotional index of the user response.

### Claim 18 (depends on 16)

The system of claim 16, wherein the virtual reality avatar is configured to be predisposed to an emotional range as defined in the first large language model input.

### Claim 19 (depends on 1)

The system of claim 1, wherein the computer-readable memory is further encoded with instructions that, when executed by the one or more processors, cause the system to:
suggest one or more rectified responses to the user, wherein the one or more rectified responses are indicative of responses that improve the final evaluation.

### Claim 20 (depends on 19)

The system of claim 19, wherein the computer-readable memory is further encoded with instructions that, when executed by the one or more processors, cause the system to:
search a repository of user data to determine whether there exists information favorable to the user for the one or more rectified responses; and
display the information favorable to the user for the one or more rectified responses alongside the one or more rectified responses.

### Claim 21

A method for training a user for a conversational encounter with a customer, the method comprising:
receiving from a user interface, a set of encounter-defining parameters, the encounter-defining parameters being descriptive of attributes of the conversational encounter;
formatting the set of encounter-defining parameters into a first large language model input, the first large language model input being suitable for processing by a large language model;
providing the large language model input to the large language model;
receiving a first large language model output from the large language model based upon the first large language model input;
providing the first large language model output to the user via the user interface;
iteratively conversing, via the user interface, with the user for a defined number of iterations, wherein iteratively conversing with the user includes the steps of:
receiving a user response via the user interface;
providing the user response to the large language model;
receiving an additional large language model output from the large language model; and
providing the additional large language model output to the user via the user interface;
providing an evaluation command to the large language model after the defined number of iterations is reached;
receiving a final evaluation from the large language model, the final evaluation being indicative of an evaluated outcome of the conversational encounter; and
displaying the final evaluation to the user via the user interface.

### Claim 22 (depends on 21)

The method of claim 21, wherein:
the first large language model input includes instructions to evaluate a direction of the user response that is indicative of the relevance of the user response; and
the direction of the user response is displayed to the user via the user interface.

### Claim 23 (depends on 21)

The method of claim 21, wherein the final evaluation includes one or more suggested improvements to the user response.

### Claim 24 (depends on 21)

The method of claim 21, wherein iteratively conversing with the user further includes:
transmitting a conversation history to the large language model, wherein the conversation history includes prior iterations of the user response and prior iterations of the additional large language model output.

### Claim 25 (depends on 21)

The method of claim 21, wherein iteratively conversing with the user further includes:
transmitting the user response to a sentiment analysis module, wherein the sentiment analysis module produces an emotional index of the user response.

### Claim 26 (depends on 25)

The method of claim 25, wherein iteratively conversing with the user further includes:
transmitting the emotional index of the user response from the sentiment analysis module to the large language model; and
instructing the large language model instruction to respond to the user response based upon the emotional index of the user response.
