Automated communications over multiple channels
A company may implement automated workflows for convenience of users or to reduce support costs. For example, allowing a user to change an address using an automated workflow may be faster or less expensive than with a human agent. Companies may provide support over different types of communications channels with different capabilities, such as voice channels and text channels. Instead of implementing different workflows for different channels, a company may separate aspects of the workflow that are common to different channels from aspects of the workflow that are different for different channels. For example, a workflow may be implemented to determine an action in response to a received communication where the action may be used with multiple channels. The action may then be used to select an action implementation that is specific to a channel.
1. A computer-implemented method for implementing an automated workflow over multiple channels, the method comprising:
creating a first communications session with a first user on a first channel using a first automated workflow, wherein the first automated workflow processes a communication of a user and selects an action from a plurality of actions;
obtaining a plurality of first channel action implementations for the first channel, wherein each first channel action implementation corresponds to an action of the plurality of actions;
receiving a first communication from the first user on the first channel;
selecting a first action of the plurality of actions by processing the first communication with the first automated workflow;
obtaining a first channel action implementation from the plurality of first channel action implementations using the first action;
transmitting a first response to the first user on the first channel using the first channel action implementation;
creating a second communications session with a second user on a second channel using the first automated workflow;
obtaining a plurality of second channel action implementations for the second channel, wherein each second channel action implementation corresponds to an action of the plurality of actions;
receiving a second communication from the second user on the second channel;
selecting the first action of the plurality of actions by processing the second communication with the first automated workflow;
obtaining a second channel action implementation from the plurality of second channel action implementations using the first action; and
transmitting a second response to the second user on the second channel using the second channel action implementation.
2. The computer-implemented method of claim 1 , wherein the first channel is a voice channel and the second channel is a customer support channel implemented on a web page.
3. The computer-implemented method of claim 1 , wherein the first channel action implementation comprises generating audio of speech to present to the first user.
4. The computer-implemented method of claim 1 , wherein selecting the first action of the plurality of actions comprises:
selecting a first intent from a plurality of intents by processing the first communication with a machine learning model; and
selecting the first action using the first intent.
5. The computer-implemented method of claim 1 , wherein selecting the first action of the plurality of actions comprises performing named entity recognition on the first communication.
6. The computer-implemented method of claim 1 , wherein selecting the first action comprises:
updating a state of the first automated workflow using the first communication; and
selecting the first action using the updated state.
7. The computer-implemented method of claim 1 , wherein selecting the first action comprises:
traversing from a first node of a directed graph to a second node of the directed graph using the first communication; and
selecting the first action using the second node of the directed graph.
8. The computer-implemented method of claim 1 , wherein:
the first response requests a first information item and a second information item from the first user; and
the second response requests the first information item and does not request the second information item from the second user.
9. A system for implementing an automated workflow over multiple channels, the system comprising:
at least one server computer comprising at least one processor and at least one memory, the at least one server computer configured to:
create a first communications session with a first user on a first channel using a first automated workflow, wherein the first automated workflow processes a communication of a user and selects an action from a plurality of actions;
obtain a plurality of first channel action implementations for the first channel, wherein each first channel action implementation corresponds to an action of the plurality of actions;
receive a first communication from the first user on the first channel;
select a first action of the plurality of actions by processing the first communication with the first automated workflow;
obtain a first channel action implementation from the plurality of first channel action implementations using the first action;
transmit a first response to the first user on the first channel using the first channel action implementation;
create a second communications session with a second user on a second channel using the first automated workflow;
obtain a plurality of second channel action implementations for the second channel, wherein each second channel action implementation corresponds to an action of the plurality of actions;
receive a second communication from the second user on the second channel;
select the first action of the plurality of actions by processing the second communication with the first automated workflow;
obtain a second channel action implementation from the plurality of second channel action implementations using the first action; and
transmit a second response to the second user on the second channel using the second channel action implementation.
10. The system of claim 9 , wherein the at least one server computer is configured to select the first action by obtaining information from the first communication.
11. The system of claim 10 , wherein the at least one server computer is configured to obtain information from the first communication by determining an intent of the first communication or performing named entity recognition on the first communication.
12. The system of claim 9 , wherein the at least one server computer is configured to:
store a first state of the first automated workflow during the first communications session to a data store;
end the first communications session;
create a third communications session with the first user on the second channel;
obtain the first state of the first automated workflow from the data store;
initialize a second instance of the first automated workflow using the first state; and
process a third communication from the first user on the second channel using the second instance of the first automated workflow.
13. The system of claim 9 , wherein the at least one server computer is configured to:
provide suggestions to users communicating on the first channel; and
not provide suggestions to users communicating on the second channel.
14. The system of claim 13 , wherein the suggestions comprise suggested messages to be sent by users or suggested completions of text being entered by users.
15. The system of claim 9 , wherein the second channel action implementation comprises presenting a form to the first user.
16. The system of claim 9 , wherein the at least one server computer is configured to obtain the first channel action implementation by using a state of the first automated workflow or a node of a directed graph.
17. One or more non-transitory, computer-readable media comprising computer-executable instructions that, when executed, cause at least one processor to perform actions comprising:
creating a first communications session with a first user on a first channel using a first automated workflow, wherein the first automated workflow processes a communication of a user and selects an action from a plurality of actions;
obtaining a plurality of first channel action implementations for the first channel, wherein each first channel action implementation corresponds to an action of the plurality of actions;
receiving a first communication from the first user on the first channel;
selecting a first action of the plurality of actions by processing the first communication with the first automated workflow;
obtaining a first channel action implementation from the plurality of first channel action implementations using the first action;
transmitting a first response to the first user on the first channel using the first channel action implementation;
creating a second communications session with a second user on a second channel using the first automated workflow;
obtaining a plurality of second channel action implementations for the second channel, wherein each second channel action implementation corresponds to an action of the plurality of actions;
receiving a second communication from the second user on the second channel;
selecting the first action of the plurality of actions by processing the second communication with the first automated workflow;
obtaining a second channel action implementation from the plurality of second channel action implementations using the first action; and
transmitting a second response to the second user on the second channel using the second channel action implementation.
18. The one or more non-transitory, computer-readable media of claim 17 , wherein the first channel action implementation comprises generating audio of speech to present to the first user and the second channel action implementation comprises presenting a form to the first user.
19. The one or more non-transitory, computer-readable media of claim 17 , wherein the plurality of first channel action implementations are specified using structured data.
20. The one or more non-transitory, computer-readable media of claim 17 , wherein the first channel action implementation comprises presenting one or more buttons to the first user.