Integration of human agent and automated tools for interactive voice response (IVR) systems

An interactive voice response system (“IVR”) comprising:

• a telephony server configured to:
• receive an initiation of a conversation submitted by a human caller using a communication network, the conversation comprising voice inputs associated with a request, the voice inputs generated by the human caller;
• provide an interface for the human caller to interact with an artificial intelligence (“AI”) engine over the communication network;
• and
• transmit to the human caller, over the communication network, responses to the voice inputs generated by the AI engine;
• and
• an application server hosting the AI engine configured to:
• receive, continuously, from the telephony server, the voice inputs generated by the human caller and the responses to the voice inputs generated by the AI engine;
• generate a machine interpretation of the voice inputs;
• monitor the voice inputs by:
• tracking a length of time of the voice inputs generated by the human caller;
• and
• monitoring the responses to the voice inputs generated by the AI engine;
• and
• wherein:
• the application server is configured to terminate the conversation when:
• the length of time is equal to a first pre-determined amount of time;
• and
• a text message from the AI agent is received at a mobile device of the human caller, the text message including a hyperlink and a selection of the hyperlink;
• and
• immediately after receipt of the text message, a delay in receiving voice inputs is detected, a length of the delay being a period of time that is equal to or greater than a second pre-determined amount of time;
• the application server is configured to transfer the human caller to a human agent when:
• the length of time is equal to the first pre-determined amount of time;
• and
• a satisfactory response to the request has not yet been provided by the AI engine;
• the application server, upon transferring, is configured to:
• transmit an instruction to the telephony server to automatically transfer the human caller to a human agent;
• link a terminal accessible by the human agent to the application server;
• transmit to the human agent, the machine interpretation of the voice inputs and the responses generated by the AI engine;
• continue the conversation, via the terminal, between the human caller and the human agent;
• simultaneous to the continuing of the conversation, continue to provide the voice inputs generated by the human caller to the AI engine;
• and
• based on the continuing conversation and the machine interpretation of the voice inputs, determine by the human agent, the request of the human caller, the request being associated with the voice inputs generated by the human caller.

A method for leveraging artificial intelligence to integrate human and machine responses within an interactive voice response (“IVR”) system, the method comprising:

• receiving an initiation of a conversation with a human caller and an artificial intelligence (“AI”) engine, the conversation comprising a request from the human caller;
• providing, to the AI engine, voice inputs generated by the human caller;
• tracking:
• sequences of utterances extracted from the voice inputs;
• and
• responses provided by the AI engine;
• automatically terminating the conversation at the IVR system when:
• the length of time is equal to a first pre-determined amount of time;
• and
• a text message from the AI agent is received at a mobile device of the human caller, the text message including a hyperlink and a selection of the hyperlink;
• and
• immediately after receipt of the text message, a delay in receiving voice inputs is detected, a length of the delay being a period of time that is equal to or greater than a second pre-determined amount of time;
• automatically transferring the human caller and the voice inputs to a human agent when:
• a sequence of utterances from the sequences of utterances is determined to be repeated more than once;
• and
• a satisfactory response to the request has not yet been provided by the AI engine, the satisfactory response comprising a response that triggers the delay of the human caller of the second pre-determined amount of time;
• continuing the conversation with the human caller and the human agent;
• based on the continuing conversation and the voice inputs, determining, by the human agent, a type of request;
• receiving a selected response provided by the human agent based on the determination of the type of request;
• transferring back the human caller to the AI engine to provide the selected response to the human caller;
• and
• wherein, when the selected response provided by the AI engine is not the satisfactory response, the method comprises:
• transmitting an alert to the human agent on a display screen positioned within proximity of the human agent;
• transferring back the human caller to the human agent;
• and
• completing the conversation via the human agent.