Methods for managing call traffic at a virtual assistant server

1. A method comprising:

receiving, by a virtual assistant server, calls which are redirected from an interactive voice response server and call metadata corresponding to the received calls provided by the interactive voice response server;

routing, by the virtual assistant server, the received calls to one of a plurality of intelligent communication modes based on routing decisions determined by a model using the call metadata, wherein each of the plurality of intelligent communication modes enable the virtual assistant server to respond to the routed calls;

determining, by the virtual assistant server, a fulfillment parameter corresponding to a caller requirement in each of the received calls;

creating, by the virtual assistant server, a transcript of each of the received calls comprising the call metadata, the routing decision, conversation data of the received call before the routing, conversational interaction between a caller of the received call and the virtual assistant server post the routing in the one of the plurality of intelligent communication modes, and the fulfillment parameter; and

training, by the virtual assistant server, the model using the transcripts.

2. The method of claim 1 , wherein the plurality of intelligent communication modes comprises chat, an intelligent agent device, voice, email, or SMS via which the virtual assistant server responds to the routed calls.

3. The method of claim 1 , further comprising outputting, by the virtual assistant server, responses to a communication received post the routing in the one of the plurality of intelligent communication modes.

4. The method of claim 1 , wherein the fulfillment parameters are determined using a natural language processing-based analysis of the transcripts.

5. The method of claim 1 , wherein the caller is a chatbot and the call metadata comprises: a unique identifier of the chatbot or a description of the chatbot.

6. The method of claim 1 , wherein the call metadata is used to determine: if the caller provided any inputs prior to the receiving, and a conversation action post the receiving.

7. The method of claim 1 , wherein a rule-based model is used to determine the routing decisions using one or more rules defined in natural language when the model is not trained.

8. A virtual assistant server comprising:

a processor; and

a memory coupled to the processor which is configured to be capable of executing programmed instructions stored in the memory to:

receive calls which are redirected from an interactive voice response server and call metadata corresponding to the received calls provided by the interactive voice response server;

route the received calls to one of a plurality of intelligent communication modes based on routing decisions determined by a model using the call metadata, wherein each of the plurality of intelligent communication modes enable the virtual assistant server to respond to the routed calls;

determine a fulfillment parameter corresponding to a caller requirement in each of the received calls;

create a transcript of each of the received calls comprising the call metadata, the routing decision, conversation data of the received call before the routing, conversational interaction between a caller of the received call and the virtual assistant server post the routing in the one of the plurality of intelligent communication modes, and the fulfillment parameter; and

train the model using the transcripts.

9. The virtual assistant server of claim 8 , wherein the plurality of intelligent communication modes comprises chat, an intelligent agent device, voice, email, or SMS via which the virtual assistant server responds to the routed calls.

10. The virtual assistant server of claim 8 , wherein the processor is further configured to be capable of executing the stored programmed instructions to:

output responses to a communication received post the routing in the one of the plurality of intelligent communication modes.

11. The virtual assistant server of claim 8 , wherein the fulfillment parameters are determined using a natural language processing-based analysis of the transcripts.

12. The virtual assistant server of claim 8 , wherein the caller is a chatbot and the call metadata comprises: a unique identifier of the chatbot or a description of the chatbot.

13. The virtual assistant server of claim 8 , wherein the call metadata is used to determine: if the caller provided any inputs prior to the receiving, and a conversation action post the receiving.

14. The virtual assistant server of claim 8 , wherein a rule-based model is used to determine the routing decisions using one or more rules defined in natural language when the model is not trained.

15. A non-transitory computer-readable medium having stored thereon instructions which when executed by a processor, causes the processor to:

receive calls which are redirected from an interactive voice response server and call metadata corresponding to the received calls provided by the interactive voice response server;

route the received calls to one of a plurality of intelligent communication modes based on routing decisions determined by a model using the call metadata, wherein each of the plurality of intelligent communication modes enable the virtual assistant server to respond to the routed calls;

determine a fulfillment parameter corresponding to a caller requirement in each of the received calls;

create a transcript of each of the received calls comprising the call metadata, the routing decisions, conversation data of the received call before the routing, conversational interaction between a caller of the received call and the virtual assistant server post the routing in the one of the plurality of intelligent communication modes, and the fulfillment parameter; and

train the model using the transcripts.

16. The non-transitory computer-readable medium of claim 15 , wherein the plurality of intelligent communication modes comprises chat, an intelligent agent device, voice, email, or SMS via which the virtual assistant server responds to the routed calls.

17. The non-transitory computer-readable medium of claim 15 , wherein the executable code, when executed by the processor further causes the processor to:

output responses to a communication received post the routing in the one of the plurality of intelligent communication modes.

18. The non-transitory computer-readable medium of claim 15 , wherein the fulfillment parameters are determined using a natural language processing-based analysis of the transcripts.

19. The non-transitory computer-readable medium of claim 15 , wherein the caller is a chatbot and the call metadata comprises: a unique identifier of the chatbot or a description of the chatbot.

20. The non-transitory computer-readable medium of claim 15 , wherein the call metadata is used to determine: if the caller provided any inputs prior to the receiving, and a conversation action post the receiving.

21. The non-transitory computer-readable medium of claim 15 , wherein a rule-based model is used to determine the routing decisions using one or more rules defined in natural language when the model is not trained.

22. A method comprising:

receiving, by a virtual assistant server, calls which are redirected from an interactive voice response server and call metadata corresponding to the received calls provided by the interactive voice response server;

for each of the received calls:

selecting, by the virtual assistant server, a predictive model from a plurality of models when training data is available;

inputting, by the virtual assistant server, the call metadata to the predictive model which determines a routing decision based on the call metadata to route the received call to one of a plurality of intelligent communication modes, wherein the routing decision is a textual output of the predictive model instructing routing the received calls to one of the plurality of intelligent communication modes;

routing, by the virtual assistant server, the received call to the one of the plurality of intelligent communication modes;

providing, by the virtual assistant server, responses to communication received post the routing in the one of the plurality of intelligent communication modes;

determining, by the virtual assistant server, a fulfillment parameter corresponding to a caller requirement in the received call; and

training, by the virtual assistant server, the predictive model using the call metadata, the routing decision, responses received post the routing decision in the one of the plurality of intelligent communication modes, and the fulfillment parameter corresponding to each of the redirected calls.

23. A method comprising:

receiving, by a virtual assistant server, calls which are redirected from an interactive voice response server and call metadata corresponding to the received calls provided by the interactive voice response server;

routing, by the virtual assistant server, the received calls to an agent device based on routing decisions made by a model using the call metadata, wherein the routing decision is a textual output of the model instructing routing the call to the agent device;

receiving, by the virtual assistant server, a fulfillment parameter for a caller requirement and textual conversation data of each of the received calls from the agent device, wherein the fulfillment parameter is determined using a natural language processing based analysis of transcripts of the received calls, and wherein the textual conversation data comprises conversational interaction between a caller of the received calls and the human agent at the agent device; and

training, by the virtual assistant server, a predictive model using the call metadata, the routing decisions, responses received post the routing decision in the one of the plurality of intelligent communication modes, the textual conversation data, and the fulfillment parameters.