UTILIZING MACHINE LEARNING MODELS TO GENERATE INTERACTIVE DIGITAL TEXT THREADS WITH PERSONALIZED AGENT ESCALATION DIGITAL TEXT REPLY OPTIONS
The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing machine learning models to determine predicted client intent classifications and/or client-agent escalation classes to generate personalized digital text reply options within an automated interactive digital text thread. For example, disclosed systems utilize the machine learning model to generate predicted client-agent escalation classes and corresponding probabilities. The disclosed systems utilize the predicted client-agent escalation classifications and the escalation class probabilities to generate personalized digital text reply options. Moreover, the disclosed systems can provide personalized digital text reply options to a client device within an automated interactive digital text thread, bypassing the inefficiency of menu options or protocols utilized to guide clients to terminal information.
1 . A computer-implemented method comprising training an agent escalation machine learning model by:
generating, utilizing the agent escalation machine learning model, a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;
identifying ground truth client-agent escalation classes generated based on interactions between agent devices and the client devices;
generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with the ground truth client-agent escalation classes; and
modifying parameters of the agent escalation machine learning model based on the measure of loss.
2 . The computer-implemented method of claim 1 , further comprising:
generating the set of training client features by extracting historical device activity of digital accounts corresponding to the client devices; and
generating, utilizing the agent escalation machine learning model, the plurality of client-agent escalation class predictions from the historical device activity of the digital accounts corresponding to the client devices.
3 . The computer-implemented method of claim 1 , wherein generating the plurality of client-agent escalation class predictions comprises utilizing at least one of a neural network or decision tree to generate the plurality of client-agent escalation class predictions.
4 . The computer-implemented method of claim 1 , wherein identifying the ground truth client-agent escalation classes comprises:
monitoring information transferred between the client devices and the agent devices; and
generating the ground truth client-agent escalation classes from the information transferred between the client devices and the agent devices.
5 . The computer-implemented method of claim 1 , wherein identifying the ground truth client-agent escalation classes comprises obtaining labeled ticket classifications based on interactions at the agent devices.
6 . The computer-implemented method of claim 1 , wherein modifying parameters of the agent escalation machine learning model based on the measure of loss comprises:
generating nodes of a decision tree model based on the measure of loss; or
modifying internal weights of a neural network based on the measure of loss.
7 . The computer-implemented method of claim 1 , further comprising training the agent escalation machine learning model based on a first layer of a hierarchical intent architecture from a plurality of layers of the hierarchical intent architecture.
8 . A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause a computer system to train an agent escalation machine learning model by:
generating, utilizing the agent escalation machine learning model, a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;
identifying ground truth client-agent escalation classes generated based on interactions between agent devices and the client devices;
generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with the ground truth client-agent escalation classes; and
modifying parameters of the agent escalation machine learning model based on the measure of loss.
9 . The non-transitory computer-readable medium of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the computer system to train the agent escalation machine learning model by:
generating the set of training client features by extracting historical device activity of digital accounts corresponding to the client devices; and
generating, utilizing the agent escalation machine learning model, the plurality of client-agent escalation class predictions from the historical device activity of the digital accounts corresponding to the client devices.
10 . The non-transitory computer-readable medium of claim 8 , wherein generating the plurality of client-agent escalation class predictions comprises utilizing at least one of a neural network or decision tree to generate the plurality of client-agent escalation class predictions.
11 . The non-transitory computer-readable medium of claim 8 , wherein identifying the ground truth client-agent escalation classes comprises:
monitoring information transferred between the client devices and the agent devices; and
generating the ground truth client-agent escalation classes from the information transferred between the client devices and the agent devices.
12 . The non-transitory computer-readable medium of claim 8 , wherein identifying the ground truth client-agent escalation classes comprises obtaining labeled ticket classifications based on interactions at the agent devices.
13 . The non-transitory computer-readable medium of claim 8 , wherein modifying parameters of the agent escalation machine learning model based on the measure of loss comprises:
generating nodes of a decision tree model based on the measure of loss; or
modifying internal weights of a neural network based on the measure of loss.
14 . The non-transitory computer-readable medium of claim 8 , further comprising instructions that, when executed by the at least one processor, cause the computer system to train the agent escalation machine learning model based on a first layer of a hierarchical intent architecture from a plurality of layers of the hierarchical intent architecture.
15 . A system comprising:
at least one processor; and
at least one non-transitory computer-readable storage medium storing instructions that, when executed by the at least one processor, cause the system to train an agent escalation machine learning model by:
generating, utilizing the agent escalation machine learning model, a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;
identifying ground truth client-agent escalation classes generated based on interactions between agent devices and the client devices;
generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with the ground truth client-agent escalation classes; and
modifying parameters of the agent escalation machine learning model based on the measure of loss.
16 . The system of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the system to train the agent escalation machine learning model by:
generating the set of training client features by extracting historical device activity of digital accounts corresponding to the client devices; and
generating, utilizing the agent escalation machine learning model, the plurality of client-agent escalation class predictions from the historical device activity of the digital accounts corresponding to the client devices.
17 . The system of claim 15 , wherein generating the plurality of client-agent escalation class predictions comprises utilizing at least one of a neural network or decision tree to generate the plurality of client-agent escalation class predictions.
18 . The system of claim 15 , wherein identifying the ground truth client-agent escalation classes comprises:
monitoring information transferred between the client devices and the agent devices; and
generating the ground truth client-agent escalation classes from the information transferred between the client devices and the agent devices.
19 . The system of claim 15 , wherein identifying the ground truth client-agent escalation classes comprises obtaining labeled ticket classifications based on interactions at the agent devices.
20 . The system of claim 15 , wherein modifying parameters of the agent escalation machine learning model based on the measure of loss comprises:
generating nodes of a decision tree model based on the measure of loss; or modifying internal weights of a neural network based on the measure of loss.