Integration of AI-powered conversational messaging with a live agent interaction

1. One or more non-transitory computer-readable media storing computer-executable instructions which, when executed by a processor on a computer system, perform a method for integrating an agent interface with AI-powered conversational messaging, the method comprising:

initiating a session at an agent device associated with a first platform, the session comprising two-way communication between a customer mobile device and the agent device;

initiating a parallel session at an interactive response system comprising a neural network and associated with a second platform, the parallel session comprising one-way communication from the customer mobile device to the interactive response system;

receiving an input from the customer mobile device at the agent device and at the interactive response system;

generating an AI-based response at the interactive response system;

displaying the response at the agent device for agent approval; and

in response to receiving agent approval, displaying the response at the customer mobile device.

2. The media of claim 1 , further comprising, at the interactive response system:

generating a prediction based on the customer input; and

based on the prediction, displaying a link to a document at the agent device.

3. The media of claim 1 , further comprising displaying the generated response on the customer mobile device without agent approval.

4. The media of claim 1 , further comprising determining an intent associated with the customer input at the interactive response system using natural language understanding.

5. The media of claim 4 , further comprising, in response to a failure to determine intent:

identifying a set of agents having a threshold approval rating;

identifying a first set of responses comprising AI-based responses generated at the interactive response system and modified by the set of agents;

identifying a second set of responses comprising AI-based responses generated at the interactive response system and rejected by the set of agents; and

mining session histories associated with the set of agents, the session histories comprising the first set of responses and the second set of responses to generate a training set for the neural network.

6. The media of claim 1 , further comprising receiving a task closure for the session at the agent device and, in response to the task closure:

generating a reminder message at the interactive response system, the reminder message generated based on the content of the session and comprising a post-session task for the agent related to the content of the session;

displaying the reminder at the agent device; and

transmitting a session history from the agent device to the interactive response system to train the neural network.

7. The media of claim 1 further comprising, at the interactive response system:

designating a topic of customer interest;

retrieving legacy communications regarding the topic;

determining whether duplicative communications are included among the legacy communications and, to the extent that duplicative communications are included in the legacy communications, removing the duplicative communications from the legacy communications;

retrieving legacy intelligence relating to historical customer selections regarding the topic;

retrieving a plurality of outcomes based on the legacy intelligence;

forming a training set for a neural network associated with the topic, the training set based on the legacy communications, legacy intelligence, and the plurality of outcomes and delimited based on an analysis of the database;

synthesizing the neural network, the neural network comprising a plurality of nodes, the synthesizing comprising using the training set to assign individual weights to each of the plurality of nodes; and

in response to a determination of intent, using the neural network to generate an AI-based response related to the topic.

8. The media of claim 7 , further comprising generating a priority score for each of a plurality of AI-generated responses to a customer input.

9. The media of claim 7 , wherein the response comprises results of legacy customer selections associated with the topic of customer interest.

10. The media of claim 7 , wherein the analysis of the database comprises:

identifying a pre-determined number of topics of interest, each topic associated with a training set; and

reducing the topics of interest found in the database to a pre-determined number of most-occurring topics of interest.

11. A system for integrating a live customer service session with AI-powered conversational messaging, the system comprising:

a mobile device comprising a mobile application configured to receive a customer input;

an interactive response system configured to:

initiate a session comprising one-way communication from the mobile device;

receive the customer input from the mobile device; and

generate an AI-based response; and

an agent device configured to:

initiate a session comprising two-way communication between the mobile device and the agent device;

display the customer input;

display the generated response from the interactive response system; and

in response to receiving agent approval, transmit the generated response to the mobile device.

12. The system of claim 11 , wherein the agent device is further configured to:

receive an agent modification of the generated response; and

in response to receiving the agent modification, transmit the generated response to the customer mobile device.

13. The system of claim 11 , the interactive response system further configured to:

generate a prediction based on the customer input; and

based on the prediction, transmit a document to the agent device.

14. The system of claim 13 , wherein agent approval of the predicted document transmits a link to the document to the mobile device.

15. The system of claim 11 , further comprising displaying the generated response at the mobile device without agent approval.

16. The system of claim 11 , wherein the interactive response system is configured to determine an intent associated with the customer input using natural language understanding.

17. The system of claim 16 , wherein, in response to a failure to determine intent, the interactive response system is configured to:

identify a set of agents having a threshold approval rating; and

mine a session history associated with the set of agents to generate a training set for the neural network.

18. The system of claim 11 , wherein:

the agent device is configured to receive a task closure; and

in response to the task closure, the interactive response system is configured to:

generate a reminder message for the agent;

transmit the reminder to the agent device; and

mine the session history from the agent device to generate a training set for the neural network.

19. A method for integrating an agent interface with AI-powered conversational messaging, the method comprising:

at a first platform comprising an agent interface:

initiating a live chat session with a customer;

receiving a customer input;

displaying an AI-based response generated at a second platform; and

in response to agent approval of the generated response, inserting the response into the live chat session; and

at the second platform comprising an interactive response system and AI engine:

receiving the customer input; and

generating an AI-based response at the interactive response system using a neural network.

20. The method of claim 19 , further comprising displaying the generated response on the first platform in an interactive window.

21. The method of claim 20 , further comprising displaying selectable options for accepting, rejecting, or modifying the generated response.