Dynamic endpoint communication channels
The present disclosure relates generally to systems, methods, and computer-readable storage media for providing a concierge service to handle a wide variety of topics and user intents via a common interface. The concierge service can be part of a connection management system that can dynamically manage and facilitate conversations between a user making a request or providing an instruction and one or more endpoints for the purposes of fulfilling the request or instruction. Such dynamic management may include transferring a communication session to a social network member endpoint based on an intent identified within natural language communications, tracking a dynamic sentiment score, and automatically switching the communication session to another endpoint based on a change in the dynamic sentiment score.
1. A non-transitory computer-readable medium comprising instructions stored thereon, the instructions executable by a connection management system to:
receive a natural language communication associated with a user device in a communication session;
interpret an intent of the natural language communication using natural language processing, wherein the natural language communication corresponds to conversational human language that does not require conformance to a rigid protocol corresponding to at least one of wake words, key words, specified phrases, or explicitly named services;
automatically determine a social network member endpoint that corresponds to the intent of the natural language communication, the social network member endpoint being selected from a collection of endpoints including at least one agent and at least one social network member;
transfer the communication session to the social network member endpoint, the communication session configured to facilitate a first conversation between the user device and the social network member endpoint, wherein the natural language communication is sent to the social network member endpoint in the communication session;
track a dynamic capability score for the communication session based on messages with a current endpoint; and
automatically switch the communication session from the current endpoint to another endpoint based on a change in the dynamic capability score.
2. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
monitor communications between the user device and the social network member endpoint in the communication session; and
detect a new intent corresponding to an action trigger associated with an action to be taken by the connection management system.
3. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
provide a message in the communication session in accordance with to a new detected intent.
4. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
send a private message to the user device, or to send a private message to the social network member endpoint in accordance with a new detected intent.
5. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
add a second endpoint to the communication session in accordance with a new detected intent; and
send a communication that is part of the communication session to the second endpoint.
6. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
drop the social network member endpoint from the communication session, and
discontinue communications that are part of the communication session from being sent to the dropped endpoint.
7. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
determine that the first conversation should be closed; and
close the first conversation with the social network member endpoint.
8. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
determine that the first conversation has gone dormant; and
send a communication as part of the first conversation to the user device to resume the first conversation.
9. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
monitor communications from the user device for a subsequent natural language communication;
interpret the intent of the subsequent natural language communication;
determine that the intent of the subsequent natural language communication does not correspond to the social network member endpoint;
determine that the intent of the subsequent natural language communication corresponds to a second endpoint; and
transfer the communication session to the second endpoint, the communication session configured to facilitate a second conversation between the user device and the second endpoint, wherein the subsequent natural language communication is sent to the second endpoint in the communication session.
10. The non-transitory computer-readable medium of claim 1 , further comprising instructions executable by the connection management system to:
continue to monitor additional communications within the communication session;
determine whether the additional communications pertain to the first conversation or a second conversation; and
forward the additional communications to the social network member endpoint when the additional communications pertain to the first conversation and forward the additional communications to a second endpoint when the additional communications pertain to the second conversation, wherein the first conversation and the second conversation are maintained in the communication session concurrently.
11. A computer-implemented method comprising:
receiving a natural language communication associated with a user device in a communication session that includes a connection management system;
interpreting an intent of the natural language communication using natural language processing, wherein the natural language communication corresponds to conversational human language that does nor require conformance to a rigid protocol corresponding to at least one of wake words, key words, specified phrases, or explicitly named services;
automatically determining a social network member endpoint that corresponds to the intent of the natural language communication, the social network member endpoint being selected from a collection of endpoints including at least one agent and at least one social network member;
transferring the communication session to the social network member endpoint, the transferred communication session configured to facilitate a first conversation between the user device and the social network member endpoint, wherein the natural language communication is sent to the social network member endpoint in the transferred communication session;
tracking a dynamic capability score for the communication session based on messages with a current endpoint; and
automatically switch the communication session from the current endpoint to another endpoint based on a change in the dynamic capability score.
12. The method of claim 11 , further comprising:
monitoring communications between the user device and the social network member endpoint in the communication session; and
detecting a new intent corresponding to an action trigger associated with an action to be taken by the connection management system.
13. The method of claim 11 , further comprising:
providing a message in the communication session in accordance with a new detected intent.
14. The method of claim 11 , further comprising:
sending a private message to the user device, or sending a private message to the social network member endpoint in accordance with a new detected intent.
15. The method of claim 11 , further comprising:
adding a second endpoint to the communication session in accordance with a new detected intent; and
sending a communication that is part of the communication session to the second endpoint.
16. The method of claim 11 , further comprising:
dropping the social network member endpoint from the communication session, and
discontinuing communications that are part of the communication session from being sent to the dropped endpoint.
17. The method of claim 11 , further comprising:
determining that the first conversation should be closed; and
closing the first conversation with the social network member endpoint.
18. The method of claim 11 , further comprising:
determining that the first conversation has gone dormant; and
sending a communication as part of the first conversation to the user device to resume the first conversation.
19. The method of claim 11 , further comprising:
monitoring communications from the user device for a subsequent natural language communication;
interpreting the intent of the subsequent natural language communication;
determining that the intent of the subsequent natural language communication does not correspond to the social network member endpoint;
determining that the intent of the subsequent natural language communication corresponds to a second endpoint; and
transferring the communication session to the second endpoint, the communication session configured to facilitate a second conversation between the user device and the second endpoint, wherein the subsequent natural language communication is sent to the second endpoint in the communication session.
20. The method of claim 11 , further comprising:
continuing to monitor additional communications within the communication session;
determine whether the additional communications pertain to the first conversation or a second conversation; and
forward the additional communications to the social network member endpoint when the additional communications pertain to the first conversation and forward the additional communications to a second endpoint when the additional communications pertain to the second conversation, wherein the first conversation and the second conversation are maintained in the communication session concurrently.
21. A system comprising:
a interface that receives a natural language communication associated with a user device in a communication session; and
a processor that executes connection management instructions, wherein the processor executes the connection management instructions to:
interprets an intent of the natural language communication using natural language processing, wherein the natural language communication corresponds to conversational human language that does not require conformance to a rigid protocol corresponding to at least one of wake words, key words, specified phrases, or explicitly named services;
automatically determines a social network member endpoint that corresponds to the intent of the natural language communication, the social network member endpoint being selected from a collection of endpoints including at least one agent and at least one social network member;
transfers the communication session to the social network member endpoint, the transferred communication session configured to facilitate a first conversation between the user device and the social network member endpoint, wherein the natural language communication is sent to the social network member endpoint in the transferred communication session;
tracks a dynamic capability score for the communication session based on messages with a current endpoint; and
automatically switches the communication session from the current endpoint to another endpoint based on a change in the dynamic capability score.
22. The system of claim 21 , the processor executing further instructions for:
monitoring communications between the user device and the social network member endpoint in the communication session; and
detecting a new intent corresponding to an action trigger associated with an action to be taken by the connection management system.
23. The system of claim 21 , the processor executing further instructions for:
providing a message in the communication session in accordance with a new detected intent.
24. The system of claim 21 , the processor executing further instructions for:
sending a private message to the user device, or sending a private message to the social network member endpoint in accordance with a new detected intent.
25. The system of claim 21 , the processor executing further instructions for:
adding a second endpoint to the communication session in accordance with a new detected intent; and
sending a communication that is part of the communication session to the second endpoint.
26. The system of claim 21 , the processor executing further instructions for:
dropping the social network member endpoint from the communication session, and
discontinuing communications that are part of the communication session from being sent to the dropped endpoint.
27. The system of claim 21 , the processor executing further instructions for:
determining that the first conversation should be closed; and
closing the first conversation with the social network member endpoint.
28. The system of claim 21 , the processor executing further instructions for:
determining that the first conversation has gone dormant; and
sending a communication as part of the first conversation to the user device to resume the first conversation.
29. The system of claim 21 , the processor executing further instructions for:
monitoring communications from the user device for a subsequent natural language communication;
interpreting the intent of the subsequent natural language communication;
determining that the intent of the subsequent natural language communication does not correspond to the social network member endpoint;
determining that the intent of the subsequent natural language communication corresponds to a second endpoint; and
transferring the communication session to the second endpoint, the communication session configured to facilitate a second conversation between the user device and the second endpoint, wherein the subsequent natural language communication is sent to the second endpoint in the communication session.
30. The system of claim 21 , the processor executing further instructions for:
continuing to monitor additional communications within the communication session;
determine whether the additional communications pertain to the first conversation or a second conversation; and
forward the additional communications to the social network member endpoint when the additional communications pertain to the first conversation and forward the additional communications to a second endpoint when the additional communications pertain to the second conversation, wherein the first conversation and the second conversation are maintained in the communication session concurrently.