Computer-implemented systems configured for automated electronic message administration and methods of use thereof
Systems and methods are provided for facilitating automated administration of email messages received by a customer service representative including a service email administrator system that reviews incoming email messages, analyzes the content, assesses emotional sentiment, urgency, and determines the best course of action. The conversation administrator is configured to use natural language processing and machine learning algorithms to generate automated responses, provide email response drafts for customer service representative (CSR) review, or route emails to CSRs based on the sentiment or urgency.
1 . A computer-implemented method for automatically routing customer messages, the method comprising:
receiving a customer message sent by a user;
analyzing the customer message using an NLP module to determine a subject of the customer message and assessing emotional state of the user by generating an emotional tone score and comparing the emotional tone score to a threshold;
assigning an urgency level based on the NLP module analysis including, when the emotional tone score exceeds the threshold, generating an alert to a customer service representative (CSR);
selecting an action in response to the urgency level, wherein: (i) for a high urgency level the action includes routing the customer message to the CSR and generating a draft response for CSR review, (ii) for a low urgency level the action includes generating an automated response, and (iii) generating the draft response or the automated response includes gathering data from at least one of a knowledge base, a subject matter base, or an external information database; and
receiving, via a user interface, an override input from the CSR that overrides an automatically determined sentiment, and training an AI model based on the override input.
2 . The method of claim 1 , wherein the customer message comprises at least one conversational-style input sent by the user.
3 . The method of claim 1 , wherein the at least one conversational-style input sent by the user comprises text data.
4 . The method of claim 1 , wherein the NLP module comprises one or more deep learning systems applied to the at least one conversational-style input to extract feature representations from the at least one conversational-style input, wherein the feature representations comprise one or more language features.
5 . The method of claim 1 , wherein the NLP module comprises one or more machine learning models to assess emotional state of the user based at least on the feature representations and generate a prediction confidence of the emotional state, wherein the one or more machine learning models comprise at least one of an or a natural language processing (NLP) model.
6 . The method of claim 1 , wherein the assessed emotional state comprises a negative state, a moderate state, or a neutral state.
7 . The method of claim 6 , wherein the urgency level comprises a high level, a medium level, or a low level.
8 . The method of claim 5 , wherein the negative emotional state corresponds to the high urgency level; the moderate emotional state corresponds to the medium urgency level; and a neutral emotional state corresponds to the low urgency level.
9 . The method of claim 8 , wherein the action comprises routing the customer message to a representative, generating a response for review, or generating an automatic response.
10 . The method of claim 9 , wherein the high urgency level corresponds to routing the customer message to a representative; the medium urgency level corresponds to generating the response for review; and the low urgency level corresponds to generating the automatic response.
11 . The method of claim 10 , wherein generating the response for review is based on the emotional state of the user.
12 . A system for automatically routing customer messages, the system comprising:
one or more computing processors; and
a machine-readable storage medium storing instructions that, when executed by the one or more processors, cause the system to:
receive a customer message sent by a user;
analyze the customer message using an NLP module to determine a subject of the customer message and assessing emotional state of the user by generating an emotional tone score and comparing the emotional tone score to a threshold;
assign an urgency level based on the NLP module analysis including, when the emotional tone score exceeds the threshold, generating an alert to a customer service representative (CSR);
select an action in response to the urgency level, wherein: (i) for a high urgency level the action includes routing the customer message to the CSR and generating a draft response for CSR review, (ii) for a low urgency level the action includes generating an automated response, and (iii) generating the draft response or the automated response includes gathering data from at least one of a knowledge base, a subject matter base, or an external information database; and
receiving, via a user interface, an override input from the CSR that overrides an automatically determined sentiment, and training an AI model based on the override input.
13 . The system of claim 12 , wherein the customer message comprises at least one conversational-style input sent by the user.
14 . The system of claim 13 , wherein the at least one conversational-style input sent by the user comprises text data.
15 . The system of claim 13 , wherein the NLP module comprises one or more deep learning systems applied to the at least one conversational-style input to extract feature representations from the at least one conversational-style input, wherein the feature representations comprise one or more language features.
16 . The system of claim 15 , wherein the NLP module comprises one or more machine learning models to assess emotional state of the user based at least on the feature representations and generate a prediction confidence of the emotional state, wherein the one or more machine learning models comprise at least one of an or a natural language processing (NLP) model.
17 . The system of claim 16 , wherein the assessed emotional state comprises a negative state, a moderate state, or a neutral state.
18 . The system of claim 17 , wherein the urgency level comprises a high level, a medium level, or a low level.
19 . The system of claim 18 , wherein the action comprises routing the customer message to a representative, generating a response for review, or generating an automatic response.
20 . The system of claim 19 , wherein the high urgency level corresponds to routing the customer message to a representative; the medium urgency level corresponds to generating the response for review; and the low urgency level corresponds to generating the automatic response.
21 . The system of claim 20 , wherein generating the response for review is based on the emotional state of the user.