Systems, media, and methods for automated response to queries made by interactive electronic chat
Systems, media, and methods for automated response to social queries comprising: monitoring queries from users, each query submitted to a vendor via an interactive chat feature of an external electronic communication platform, monitoring human responses to the queries, monitoring subsequent communications conducted via the electronic communication platform until each query is resolved; applying a first machine learning algorithm to the monitored communications to identify a query susceptible to response automation; applying a second machine learning algorithm to the query susceptible to response automation to identify one or more responses likely to resolve the query; and either i) notifying a human to respond to the query susceptible to response automation with the one or more responses likely to resolve the query, or ii) instantiating an autonomous software agent configured to respond to the query susceptible to response automation with the one or more responses likely to resolve the query.
1. A computer-implemented method, comprising:
monitoring, by a computing device, an enterprise system by using an API to interact with enterprise data corresponding to user interactions with the enterprise system, wherein the user interactions have corresponding user interaction data, wherein the computing device includes a query-detection machine-learning model and a resolution machine-learning model, and wherein the query-detection machine-learning model is a distinct machine-learning model from the query-detection machine-learning model;
integrating the enterprise data and the user interaction data into a query-detection machine-learning model;
detecting input to the enterprise system;
identifying a query in the input using the query-detection machine-learning model and the input;
executing a resolution machine-learning model using the query, wherein the resolution machine-learning model identifies one or more responses that are likely to resolve the query, wherein the one or more responses are associated with a relevance value assigned based on historical queries and responses;
automatically responding to the query with a particular response that is likely to resolve the query from the one or more responses, wherein the particular response is selected based on the relevance value; and
instantiating an autonomous software agent configured to provide the particular response as a response to other queries.
2. The computer-implemented method of claim 1 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include a decision tree that is a predictive model, and wherein the response is determined using one or more input variables associated with the query.
3. The computer-implemented method of claim 1 , wherein the input is parsed using natural language.
4. The computer-implemented method of claim 1 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include an algorithm that is trained using prior queries and responses.
5. The computer-implemented method of claim 1 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include machine logic that identifies terms, and wherein the identified terms are used to generate a narrative flow.
6. The computer-implemented method of claim 1 , wherein the query-detection machine-learning model and/or the resolution machine-learning model use an API to facilitate communication between two systems.
7. The computer-implemented method of claim 1 , wherein the query-detection machine-learning model and/or the resolution machine-learning model determine the one or more responses to a query using one or more input variables from the query.
8. The computer-implemented method of claim 1 , wherein executing the resolution machine-learning model further comprises:
assigning confidence values to the one or more responses; and
selecting, from the one or more responses, a particular response that is likely to resolve the query based on the confidence values.
9. A system, comprising:
one or more data processors; and
a non-transitory computer readable storage medium containing instructions which when executed on the one or more data processors, cause the one or more data processors to perform operations including:
monitoring an enterprise system by using an API to interact with enterprise data corresponding to user interactions with the enterprise system, wherein the user interactions have corresponding user interaction data, wherein the system includes a query-detection machine-learning model and a resolution machine-learning model, and wherein the query-detection machine-learning model is a distinct machine-learning model from the query-detection machine-learning model;
integrating the enterprise data and the user interaction data into a response query-detection machine-learning model;
detecting input to the enterprise system;
identifying a query in the input using the query-detection machine-learning model and the input;
executing a resolution machine-learning model using the query, wherein the resolution machine-learning model identifies one or more responses that are likely to resolve the query, wherein the one or more responses are associated with a relevance value assigned based on historical queries and responses;
automatically responding to the query with a particular response of the one or more responses, wherein the particular response is selected based on the relevance value; and
instantiating an autonomous software agent configured to provide the particular response as a response to other queries.
10. The system of claim 9 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include a decision tree that is a predictive model, and wherein the response is determined using one or more input variables associated with the query.
11. The system of claim 9 , wherein the input is parsed using natural language.
12. The system of claim 9 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include an algorithm that is trained using prior queries and responses.
13. The system of claim 9 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include machine logic that identifies terms, and wherein the identified terms are used to generate a narrative flow.
14. The system of claim 9 , wherein the query-detection machine-learning model and/or the resolution machine-learning model use an API to facilitate communication between two systems.
15. The system of claim 9 , wherein the query-detection machine-learning model and/or the resolution machine-learning model determine the one or more responses to a query using one or more input variables from the query.
16. The system of claim 9 , wherein executing the resolution machine-learning model further comprises:
assigning confidence values to the one or more responses; and
selecting, from the one or more responses, a particular response that is likely to resolve the query based on the confidence values.
17. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause a data processing apparatus to perform operations including:
monitoring an enterprise system by using an API to interact with enterprise data corresponding to user interactions with the enterprise system, wherein the user interactions have corresponding user interaction data, wherein the non-transitory machine-readable storage medium includes a query-detection machine-learning model and a resolution machine-learning model, and wherein the query-detection machine-learning model is a distinct machine-learning model from the query-detection machine-learning model;
integrating the enterprise data and the user interaction data into a query-detection machine-learning model;
detecting input to the enterprise system;
identifying a query in the input using the query-detection machine-learning model and the input;
executing a resolution machine-learning model using the query, wherein the resolution machine-learning model identifies one or more responses that are likely to resolve the query, wherein the one or more responses are associated with a relevance value assigned based on historical queries and responses;
automatically responding to the query with a particular response of the one or more responses, wherein the particular response is selected based on the relevance value; and
instantiating an autonomous software agent configured to provide the particular response as a response to other queries.
18. The computer-program product of claim 17 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include a decision tree that is a predictive model, and wherein the response is determined using one or more input variables associated with the query.
19. The computer-program product of claim 17 , wherein the input is parsed using natural language.
20. The computer-program product of claim 17 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include an algorithm that is trained using prior queries and responses.
21. The computer-program product of claim 17 , wherein the query-detection machine-learning model and/or the resolution machine-learning model include machine logic that identifies terms, and wherein the identified terms are used to generate a narrative flow.
22. The computer-program product of claim 17 , wherein the query-detection machine-learning model and/or the resolution machine-learning model use an API to facilitate communication between two systems.
23. The computer-program product of claim 17 , wherein the query-detection machine-learning model and/or the resolution machine-learning model determine the one or more responses to a query using one or more input variables from the query.
24. The computer-program product of claim 17 , wherein executing the resolution machine-learning model further comprises:
assigning confidence values to the one or more responses; and
selecting, from the one or more responses, a particular response that is likely to resolve the query based on the confidence values.