Systems and methods relating to generating simulated interactions for training contact center agents

Patent metadata

Field	Value
Publication number	US-12598254-B2
Application number	US-202318129192-A
Country	US
Kind code	B2
Filing date	Mar 31, 2023
Priority date	Feb 16, 2023
Publication date	Apr 7, 2026
Grant date	Apr 7, 2026

Abstract

A method for generating a customer bot and using the customer bot to train agents, where a first process generates the customer bot and a second process uses the customer bot to train the agents. The first process includes: gathering conversation data; mining intents from the conversation data; constructing, from the mined intents, a dialog engine simulating an interaction type; uploading the customer bot to an automated training module for use thereby; and periodically repeating the previous steps so to update the customer bot with recent conversation data. The second process includes: monitoring for triggering events; initiating the training by initiating a virtual communication to a user device of a first agent; connecting the virtual communication to the customer bot; conducting a simulated interaction; and analyzing one or more statements received from the first agent to derive a performance assessment.

First claim

That which is claimed:

1 . A system for generating a customer bot and using the customer bot to train agents in a contact center, the system comprising:
a bot generating module; an automated training module; at least one processor; and at least one memory comprising a plurality of instructions stored therein that, in response to execution by the at least one processor, causes the system to perform, via the bot generating module, a first process that generates the customer bot and a second process, via an automated training module, that uses the customer bot to train the agents; wherein the first process comprises the steps of:
gathering conversation data, the conversation data comprising data derived from natural language conversations occurring in the contact center during interactions between the agents and customers;
mining intents from the conversation data, wherein each mined intent comprises an intent label and a set of utterances associated with the intent label, the intent label identifying an issue found to be recurring within the interactions of the conversation data, and the set of utterances comprising utterances used to raise, discuss, or resolve the issue, wherein the utterances of the set of utterances comprise statements or phrases made by both the customers and agents;
selecting one or more mined intents of the mined intents, the one or more mined intents selected based on a relatedness to an interaction type found within the interactions;
constructing, from sets of utterances of the selected one or more mined intents, a dialog engine simulating the interaction type, the dialog engine defining a dialog flow for navigating the one or more issues associated with the selected one or more mined intents, the dialog flow including both customer-side statements and agent-side statements;
generating the customer bot with the dialog engine, the customer bot being configured in accordance with the customer-side statements of the dialog flow so to mimic a customer;
uploading the customer bot to the automated training module for use thereby to train the agents pursuant to the second process; and
periodically repeating the previous steps of the first process so that the customer bot uploaded to the automated training module is updated with conversation data that has been gathered since a previous time of repeating the steps of the first process;
wherein the second process comprises the steps of:
monitoring for one or more triggering events that determine whether a first agent of the agents should receive training related to the interaction type;
in response to detecting the one or more triggering events, initiating the training by initiating a virtual communication to a user device of the first agent;
connecting the virtual communication to the customer bot in response to establishing a communication connection with the user device;
conducting a simulated interaction of the interaction type by determining a scenario as a function of an evaluation of a performance of the first agent with respect to the interaction type in one or more previous interactions prior to the determination of the scenario and transmitting one or more customer-statements generated by the customer bot in accordance with the determined scenario to the first agent and receiving one or more statements made by the first agent in response thereto;
analyzing the one or more statements received from the first agent to derive a performance assessment of the first agent.

2 . The system of claim 1 , wherein the second process further comprises generating an electronic communication of the performance assessment and then transmitting the generated electronic communication to a predetermined user device, the predetermined user device comprising at least one of the user device of the first agent or a user device of a manager of the first agent.

3 . The system of claim 1 , wherein the natural language conversations comprise at least one of chats and voice calls.

4 . The system of claim 1 , wherein the natural language conversations comprise both chats and voice calls; and
wherein, for the voice calls, the first process further comprises transcribing each of the voice calls into text via speech recognition.

5 . The system of claim 1 , wherein the step of analyzing the one or more statements received from the first agent to derive the performance evaluation for the first agent comprises comparing the one or more statements received from the first agent against comparable ones of the agent-statements included in the dialog flow.

6 . The system of claim 1 , wherein the step of analyzing the one or more statements received from the first agent to derive the performance evaluation for the first agent comprises key word detection to determine a completeness of the one or more statements received from the first agent.

7 . The system of claim 1 , wherein the step of analyzing the one or more statements received from the first agent to derive the performance evaluation for the first agent comprises speech and text analytics that includes:
topic detection analysis; and positive and negative sentiment analysis.

8 . The system of claim 1 , wherein the one or more triggering events comprises:
a determination that the user device if the first agent is logged in and able to receive the virtual communication and the first agent is available to conduct the simulated interaction; and receiving input from a predetermined user device that selects the first agent for receiving the training related to the simulated interaction, the predetermined user device comprising at least one of the user device of the first agent or a user device of a manager of the first agent.

9 . The system of claim 1 , wherein the one or more triggering events comprise receiving performance metrics measuring a performance of the first agent in interactions of a same type as the interaction type with customers of the contact center.

10 . The system of claim 1 , wherein the virtual communication comprises a voice call; and
wherein the second process further comprises generating audio for the customer-side statements via a text-to-speech conversion.

11 . The system of claim 1 , wherein to mine intents from the conversation data comprises applying an intent mining algorithm that includes:
wherein the intent mining algorithm comprises:
analyzing utterances occurring in the conversation data to identify intent-bearing utterances, wherein an intent-bearing utterance is defined as one of the utterances determined to have an increased likelihood of expressing an intent;
analyzing the identified intent-bearing utterances to identify candidate intents, wherein the candidate intents are each identified as being a text phrase occurring within one of the intent-bearing utterances that has two parts: an action, which comprises a word or phrase describing a purpose or task, and an object, which comprises a word or phrase describing an object or thing upon which the action operates;
selecting, in accordance with one or more criteria, salient intents from the candidate intents;
grouping the selected salient intents into salient intent groups in accordance with a degree of semantic similarity between the salient intents;
for each of the salient intent groups, selecting one of the salient intents as the intent label and designating the other of the salient intents as the intent alternatives; and
associating the intent-bearing utterances with the salient intent groups via determining a degree of semantic similarity between: the candidate intents present in the intent-bearing utterance; and the intent alternatives within each of the salient intent

Assignees

• Genesys Cloud Services Inc

Inventors

• Ferris Therese M
• Coleman Travis Dale
• Danish Mathew Scott
• Mouslmani Taleb John

Classifications

• H04M2203/403

  Agent or workforce training · CPC title

• H04M2201/40

  using speech recognition · CPC title

• H04M2201/39

  using speech synthesis · CPC title

• H04M3/5191

  interacting with the Internet · CPC title

• H04L51/02

  using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages · CPC title