Method and system to manage tech support interactions using dynamic notification platform
In general, embodiments relate to a method for managing a technical support (TS) session on a technical support system. The embodiments include receiving TS correspondence from a client, wherein the TS correspondence is associated with the TS session, classifying the TS correspondence to assign it a question classification, based on the question classification, making a determination that the TS correspondence matches at least one prior received TS correspondence, wherein the at least one prior received TS correspondence is associated with the TS session, and visually identifying the TS correspondence and the at least one prior received TS correspondence on a graphical customer interface (GUI) of the technical support system.
1. A method for managing a technical support (TS) session on a technical support system, the method comprising:
receiving TS correspondence from a client, wherein the TS correspondence is received from the client during a TS session;
classifying, by a classifier, the TS correspondence to assign it a question classification, wherein the classifier classifies to the question classification, a statement classification, or a command classification using a natural language processing model;
storing the TS correspondence in a question storage when the TS correspondence has the question classification;
based on the question classification, making a determination that the TS correspondence received during the TS session matches at least one prior received TS correspondence, wherein the at least one prior received TS correspondence is associated with the same TS session;
visually identifying the TS correspondence and the at least one prior received TS correspondence from the client on a graphical customer interface (GUI) of a technical support system, wherein the TS correspondence and the at least one prior received TS correspondence are highlighted to distinguish from other TS correspondences displayed in the GUI that do not match the at least one prior received TS correspondence;
receiving a second TS correspondence from the client, wherein the second TS correspondence is associated with the TS session;
classifying the second TS correspondence to assign it to a statement classification; and
based on the statement classification, not storing the second TS correspondence in the question storage, wherein the question storage is used to train the classifier.
2. The method of claim 1 , wherein the TS correspondence comprises text based on audio input obtained from the client.
3. The method of claim 1 , wherein the TS correspondence comprises text obtained via a chat session from the client.
4. The method of claim 3 , wherein prior to classifying the TS correspondence, the text obtained via the chat session is cleaned to obtain cleaned TS correspondence, wherein classifying the TS correspondence uses the cleaned TS correspondence.
5. The method of claim 4 , wherein cleaning the TS correspondence comprises modifying at least a portion of the text to address at least one grammatical error.
6. A technical support (TS) system, comprising:
a processor;
wherein the TS system is configured to:
receive TS correspondence from a client, wherein the TS correspondence is received from the client during a TS session;
classify, by a classifier, the TS correspondence to assign it a question classification, wherein the classifier classifies to the question classification, a statement classification, or a command classification using a natural language processing model;
store the TS correspondence in a question storage when the TS correspondence has the question classification;
based on the question classification, make a determination that the TS correspondence received during the TS session matches at least one prior received TS correspondence, wherein the at least one prior received TS correspondence is associated with the same TS session; and
visually identify the TS correspondence and the at least one prior received TS correspondence from the client on a graphical customer interface (GUI) of a technical support system, wherein the TS correspondence and the at least one prior received TS correspondence are highlighted to distinguish from other TS correspondences displayed in the GUI that do not match the at least one prior received TS correspondence
receive a second TS correspondence from the client, wherein the second TS correspondence is associated with the TS session;
classify the second TS correspondence to assign it to a statement classification; and
based on the statement classification, not store the second TS correspondence in the question storage, wherein the question storage is used to train the classifier.
7. The technical support system of claim 6 , wherein the TS correspondence comprises text based on audio input obtained from the client.
8. The technical support system of claim 6 , wherein the TS correspondence comprises text obtained via a chat session from the client.
9. The technical support system of claim 8 , wherein prior to classifying the TS correspondence, the text obtained via the chat session is cleaned to obtain cleaned TS correspondence, wherein classifying the TS correspondence uses the cleaned TS correspondence.
10. The technical support system of claim 9 , wherein cleaning the TS correspondence comprises modifying at least a portion of the text to address at least one grammatical error.
11. A non-transitory computer readable medium comprising computer readable program code to:
receive TS correspondence from a client, wherein the TS correspondence is received from the client during a TS session;
classify, by a classifier, the TS correspondence to assign it a question classification, wherein the classifier classifies to the question classification, a statement classification, or a command classification using a natural language processing model;
store the TS correspondence in a question storage when the TS correspondence has the question classification;
based on the question classification, make a determination that the TS correspondence received during the TS session matches at least one prior received TS correspondence, wherein the at least one prior received TS correspondence is associated with the same TS session; and
visually identify the TS correspondence and the at least one prior received TS correspondence from the client on a graphical customer interface (GUI) of a technical support system, wherein the TS correspondence and the at least one prior received TS correspondence are highlighted to distinguish from other TS correspondences displayed in the GUI that do not match the at least one prior received TS correspondence;
receive a second TS correspondence from the client, wherein the second TS correspondence is associated with the TS session;
classify the second TS correspondence to assign it to a statement classification; and
based on the statement classification, not store the second TS correspondence in the question storage, wherein the question storage is used to train the classifier.
12. The non-transitory computer readable medium of claim 11 , wherein the TS correspondence comprises text based on audio input obtained from the client.
13. The non-transitory computer readable medium of claim 11 , wherein the TS correspondence comprises text obtained via a chat session from the client.
14. The non-transitory computer readable medium of claim 13 ,
wherein prior to classifying the TS correspondence, the text obtained via the chat session is cleaned to obtain cleaned TS correspondence, wherein classifying the TS correspondence uses the cleaned TS correspondence; and
wherein cleaning the TS correspondence comprises modifying at least a portion of the text to address at least one grammatical error.