Utilizing machine learning models to generate interactive digital text threads with personalized agent escalation digital text reply options

1. A computer-implemented method comprising:

training an agent escalation machine learning model by:

generating a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;

generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with ground truth client-agent escalation classes generated from agent devices that interacted with the client devices; and

modifying parameters of the agent escalation machine learning model based on the measure of loss;

extracting client features corresponding to a client device participating in an automated interactive digital text thread;

generating, utilizing the agent escalation machine learning model, a plurality of predicted client-agent escalation classes and a plurality of escalation class probabilities from the client features;

selecting at least two predicted client-agent escalation classes from the plurality of predicted client-agent escalation classes utilizing the plurality of escalation class probabilities; and

providing, for display via the client device, at least two personalized escalation digital text reply options corresponding to the at least two predicted client-agent escalation classes via the automated interactive digital text thread.

2. The computer-implemented method of claim 1 , further comprising generating, utilizing an intent machine learning model different from the agent escalation machine learning model, a plurality of predicted client intent classifications different from the plurality of predicted client-agent escalation classes.

3. The computer-implemented method of claim 1 , further comprising, in response to user interaction with a personalized digital text reply option of the at least two personalized escalation digital text reply options corresponding to a predicted client-agent escalation class, initiating a client self-service workflow corresponding to the predicted client-agent escalation class.

4. The computer-implemented method of claim 1 , further comprising:

providing a selectable escalation option within the automated interactive digital text thread; and

in response to detecting a user interaction with the selectable escalation option, providing the at least two personalized escalation digital text reply options corresponding to the at least two predicted client-agent escalation classes for display.

5. The computer-implemented method of claim 2 , further comprising:

providing, for display via the client device, at least two personalized intent digital classification options corresponding to at least two predicted client intent classifications of the plurality of predicted client intent classifications via the automated interactive digital text thread.

6. The computer-implemented method of claim 1 , wherein utilizing the agent escalation machine learning model comprises generating the plurality of predicted client-agent escalation classes and the plurality of escalation class probabilities utilizing one or more of a random forest model or gradient boosted decision tree model.

7. The computer-implemented method of claim 1 , wherein extracting client features comprises one or more of determining a digital account value, a direct deposit status of a digital account, or application device activity on the digital account.

8. The computer-implemented method of claim 1 , further comprising:

in response to detecting an escalation interaction via the automated interactive digital text thread, initiating a client-agent response session between the client device and an agent device; and

providing, for display, one or more of the plurality of predicted client-agent escalation classes to the agent device.

9. 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 a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;

generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with ground truth client-agent escalation classes generated from agent devices that interacted with the client devices; and

modifying parameters of the agent escalation machine learning model based on the measure of loss;

extract client features corresponding to a client device participating in an automated interactive digital text thread;

generate, utilizing the agent escalation machine learning model, a plurality of predicted client-agent escalation classes and a plurality of escalation class probabilities from the client features;

select at least two predicted client-agent escalation classes from the plurality of predicted client-agent escalation classes utilizing the plurality of escalation class probabilities; and

provide, for display via the client device, at least two personalized escalation digital text reply options corresponding to the at least two predicted client-agent escalation classes via the automated interactive digital text thread.

10. The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the at least one processor, further cause the computer system to generate, utilizing an intent machine learning model different from the agent escalation machine learning model, a plurality of predicted client intent classifications different from the plurality of predicted client-agent escalation classes.

11. The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the at least one processor, further cause the computer system to, in response to user interaction with a personalized digital text reply option of the at least two personalized escalation digital text reply options corresponding to a predicted client-agent escalation class, in initiate a client self-service workflow corresponding to the predicted client-agent escalation class.

12. The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the at least one processor, further cause the computer system to:

provide a selectable escalation option within the automated interactive digital text thread; and

in response to detecting a user interaction with the selectable escalation option, provide the at least two personalized escalation digital text reply options corresponding to the at least two predicted client-agent escalation classes for display.

13. The non-transitory computer-readable medium of claim 10 , wherein the instructions, when executed by the at least one processor, further cause the computer system to:

provide, for display via the client device, at least two personalized intent digital classification options corresponding to at least two predicted client intent classifications of the plurality of predicted client intent classifications via the automated interactive digital text thread.

14. The non-transitory computer-readable medium of claim 9 , wherein the instructions, when executed by the at least one processor, further cause the computer system to:

in response to detecting an escalation interaction via the automated interactive digital text thread, initiate a client-agent response session between the client device and an agent device; and

provide, for display, one or more of the plurality of predicted client-agent escalation classes to the agent device.

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 a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;

generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with ground truth client-agent escalation classes generated from agent devices that interacted with the client devices; and

modifying parameters of the agent escalation machine learning model based on the measure of loss;

extract client features corresponding to a client device participating in an automated interactive digital text thread;

generate, utilizing the agent escalation machine learning model, a plurality of predicted client-agent escalation classes and a plurality of escalation class probabilities from the client features;

select at least two predicted client-agent escalation classes from the plurality of predicted client-agent escalation classes utilizing the plurality of escalation class probabilities; and

provide, for display via the client device, at least two personalized escalation digital text reply options corresponding to the at least two predicted client-agent escalation classes via the automated interactive digital text thread.

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 a plurality of client-agent escalation class predictions from a set of training client features corresponding to client devices;

generating a measure of loss by comparing, utilizing a loss function, the plurality of client-agent escalation class predictions with ground truth client-agent escalation classes generated from agent devices that interacted with the client devices; and

modifying parameters of the agent escalation machine learning model based on the measure of loss.

17. The system of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the system to, in response to user interaction with a personalized digital text reply option of the at least two personalized escalation digital text reply options corresponding to a predicted client-agent escalation class, in initiate a client self-service workflow corresponding to the predicted client-agent escalation class.

18. The system of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the system to:

provide a selectable escalation option within the automated interactive digital text thread; and

in response to detecting a user interaction with the selectable escalation option, provide the at least two personalized escalation digital text reply options corresponding to the at least two predicted client-agent escalation classes for display.

19. The system of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the system to:

generate, utilizing an intent machine learning model different from the agent escalation machine learning model, a plurality of predicted client intent classifications different from the plurality of predicted client-agent escalation classes; and

provide, for display via the client device, at least two personalized intent digital classification options corresponding to at least two predicted client intent classifications of the plurality of predicted client intent classifications via the automated interactive digital text thread.

20. The system of claim 15 , further comprising instructions that, when executed by the at least one processor, cause the system to:

in response to detecting an escalation interaction via the automated interactive digital text thread, initiate a client-agent response session between the client device and an agent device; and

provide, for display, one or more of the plurality of predicted client-agent escalation classes to the agent device.