Method and apparatus for facilitating agent conversations with customers of an enterprise

1. A computer-implemented method for facilitating agent conversations with customers of an enterprise, the method comprising:

receiving, by a processor, a request for a conversation with an agent, the request provided by a customer;

connecting the customer to a human agent by the processor in response to the receipt of the request for the conversation;

for an input provided by the customer during the conversation with the human agent, determining by the processor, whether a Virtual Agent (VA) is capable of providing a faster response to the input of the customer than the human agent; and

deflecting the conversation, by the processor, from the human agent to the VA to respond to the input of the customer based at least in part on the determination that the VA is capable of providing the faster response to the input of the customer than the human agent,

wherein the conversation is deflected back from the VA to the human agent for a subsequent input if it is determined that the VA is not capable of providing the faster response to the subsequent input of the customer than the human agent.

2. The method as claimed in claim 1 , wherein providing the faster response to the input comprises fetching information from one of a database and a customer relationship management (CRM) system.

3. The method as claimed in claim 1 , wherein the conversation is deflected from the human agent to the VA if the response is readily available with the VA on account of tracking of the conversation by the VA.

4. The method as claimed in claim 1 , further comprising:

predicting, by the processor, one or more intents corresponding to the input provided by the customer;

computing, by the processor, a confidence score corresponding to each intent from among the one or more intents, the confidence score indicative of an ability of the VA to provide an effective response to the input provided by the customer; and

comparing, by the processor, the confidence score corresponding to each intent with a predefined threshold score,

wherein the conversation is deflected from the human agent to the VA if the confidence score of at least one intent from among the one or more intents is greater than or equal to the predefined threshold score and the VA is capable of providing the faster response to the input of the customer than the human agent.

5. The method as claimed in claim 4 , wherein the one or more intents corresponding to the input are predicted based on at least one of the input, current conversation history and a past conversation history of the customer.

6. The method as claimed in claim 1 , further comprising:

passing, by the processor, a contextual information related to the conversation between the customer and the human agent to the VA subsequent to the deflection of the conversation from the human agent to the VA.

7. The method as claimed in claim 1 , wherein the conversation is deflected to one or more VAs during a course of the conversation based at least in part on a capability of respective VA from among the two or more VAs to provide faster effective response to a corresponding customer input.

8. The method as claimed in claim 1 , further comprising:

maintaining, by the processor, a log of deflections between the human agent and one or more VAs for each customer conversation handled by the human agent; and

training the human agent in customer conversation handling at periodic intervals by the processor based on the log of deflections.

9. An apparatus for facilitating agent conversations with customers of an enterprise, the apparatus comprising:

a memory for storing instructions; and

a processor configured to execute the instructions and thereby cause the apparatus to at least:

receive a request for a conversation with an agent, the request provided by a customer;

connect the customer to a human agent in response to the receipt of the request for the conversation;

for an input provided by the customer during the conversation with the human agent, determine whether a Virtual Agent (VA) is capable of providing a faster response to the input of the customer than the human agent; and

deflect the conversation from the human agent to the VA to respond to the input of the customer based at least in part on the determination that the VA is capable of providing the faster response to the input of the customer than the human agent,

wherein the conversation is deflected back from the VA to the human agent for a subsequent input if it is determined that the VA is not capable of providing the faster response to the subsequent input of the customer than the human agent.

10. The apparatus as claimed in claim 9 , wherein providing the faster response to the input comprises fetching information from one of a database and a customer relationship management (CRM) system.

11. The apparatus as claimed in claim 9 , wherein the conversation is deflected from the human agent to the VA if the response is readily available with the VA on account of tracking of the conversation by the VA.

12. The apparatus as claimed in claim 11 , wherein the apparatus is further caused to:

predict one or more intents corresponding to the input provided by the customer;

compute a confidence score corresponding to each intent from among the one or more intents, the confidence score indicative of an ability of the VA to provide an effective response to the input provided by the customer; and

compare the confidence score corresponding to each intent with a predefined threshold score,

wherein the conversation is deflected from the human agent to the VA if the confidence score of at least one intent from among the one or more intents is greater than or equal to the predefined threshold score and the VA is capable of providing the faster response to the input of the customer than the human agent.

13. The apparatus as claimed in claim 12 , wherein the one or more intents corresponding to the input are predicted based on at least one of the input, current conversation history and a past conversation history of the customer.

14. The apparatus as claimed in claim 9 , wherein the apparatus is further caused to:

pass a contextual information related to the conversation between the customer and the human agent to the VA subsequent to the deflection of the conversation from the human agent to the VA.

15. The apparatus as claimed in claim 9 , wherein the conversation is deflected to one or more VAs during a course of the conversation based at least in part on a capability of respective VA from among the two or more VAs to provide faster effective response to a corresponding customer input.

16. The apparatus as claimed in claim 9 , wherein the apparatus is further caused to:

maintain a log of deflections between the human agent and one or more VAs for each customer conversation handled by the human agent; and

train the human agent in customer conversation handling at periodic intervals by the processor based on the log of deflections.

17. A computer-implemented method for facilitating agent conversations with customers of an enterprise, the method comprising:

receiving, by a processor, a request for a conversation with an agent, the request provided by a customer;

connecting the customer to a human agent by the processor in response to the receipt of the request for the conversation;

for an input provided by the customer during the conversation with the human agent, predicting, by the processor, one or more intents corresponding to the input provided by the customer;

computing, by the processor, a confidence score corresponding to each intent from among the one or more intents, the confidence score indicative of an ability of a Virtual Agent (VA) to provide an effective response to the input provided by the customer;

comparing, by the processor, the confidence score corresponding to each intent with a predefined threshold score;

determining by the processor, whether the VA is capable of providing a faster response to the input of the customer than the human agent; and

deflecting the conversation, by the processor, from the human agent to the VA to respond to the input of the customer based at least in part on the determination that the confidence score of at least one intent from among the one or more intents is greater than or equal to the predefined threshold score and the VA is capable of providing the faster response to the input of the customer than the human agent,

wherein the conversation is deflected back from the VA to the human agent for a subsequent input if it is determined that the VA is not capable of providing the faster response to the subsequent input of the customer than the human agent.

18. The method as claimed in claim 17 , wherein providing the faster response to the input comprises fetching information from one of a database and a customer relationship management (CRM) system.

19. The method as claimed in claim 17 , wherein the conversation is deflected to one or more VAs during a course of the conversation based at least in part on a capability of respective VA from among the two or more VAs to provide faster effective response to a corresponding customer input.

20. The method as claimed in claim 17 , further comprising:

maintaining, by the processor, a log of deflections between the human agent and one or more VAs for each customer conversation handled by the human agent; and

training the human agent in customer conversation handling at periodic intervals by the processor based on the log of deflections.