System and method for managing customer call-backs
System and method for automatically calling back a customer via a predictive model determines a plurality of call-back metrics for a plurality of advisor records. The predictive model is applied to call-back data to identify customers that are likely to require a series of call-backs, and automatically generates a preferred call-back to such customers to reduce this risk. The automated call-back may follow termination of an identified customer's inbound call, or at some time after completion of a previous call interaction of the identified customer with an advisor. In the predictive model, a first compilation of call-back metrics record is representative of an overall likelihood of call-backs associated with each advisor record, and a second compilation of the plurality of call-back metrics is representative of a likelihood of call-backs for each of the plurality of products of the enterprise associated with the advisor record.
1 . A method comprising:
training, by a processor, a predictive machine learning model using call-back metrics of a set of callers of a set of historical calls, a set of agent records of a set of agents of a call center that participated in the set of historical calls, and time series forecasting, wherein training the predictive machine learning model includes applying regularization for a plurality of iterations;
retrieving, by the processor, data associated with a caller of an inbound call of the call center between the caller and an agent of the call center, the data comprising at least an interaction of the caller with an interactive voice response unit (IVR) during the inbound call and an agent record associated with the agent of the call center;
in response to determining that the inbound call has terminated:
executing, by the processor, a predictive machine learning model, using the data including the interaction with the IVR during the inbound call and the agent record of the agent, to generate output data comprising (i) a call-back signal corresponding to a likelihood of repeated call-backs by the caller, and (ii) an interaction signal representative of a likelihood that the caller will perform a predetermined interaction during a future call-back, as a result of the termination of the inbound call between the caller and the agent;
classifying, by the predictive machine learning model executing via the processor based on the call-back signal determined by the predictive machine learning model, the caller into one of a first call-back group and a second call-back group;
in a first event where the predictive machine learning model classifies the caller into the first call-back group, instructing, by the processor, an automatic calling device in communication with the processor to automatically initiate an electronic communication session between the caller and at least one agent of the call center; and
in a second event where the predictive machine learning model classifies the caller into the second call-back group, instructing, by the processor, the automatic calling device in communication with the processor to automatically execute a subordinate call-back procedure.
2 . The method of claim 1 , wherein the inbound call is terminated based on the caller exercising a call-back option or by abandoning the inbound call.
3 . The method of claim 1 , wherein the data associated with the caller comprises at least one of an interaction of the caller with at least one advisor of an enterprise associated with the call center or call-back data associated with the caller.
4 . The method of claim 1 , wherein the electronic communication session is initiated between the caller and a preferred agent of the call center.
5 . The method of claim 1 , wherein the call-back signal further corresponds to a likelihood that the caller will accept an offer to purchase one of a plurality of products.
6 . The method of claim 1 , wherein the call-back signal further corresponds to a likelihood that the caller will lapse in payments for a purchased product.
7 . The method of claim 1 , wherein the subordinate call-back procedure comprises one or more of automatically calling back with a recorded message of the IVR, automatically calling back for connection to an agent from a pool of unskilled agents, and automatically deferring call-back.
8 . A system for managing customer calls within a call center, the system comprising:
an automatic calling device; and
a processor, in communication with the automatic calling device, the processor configured to:
train a predictive machine learning model using call-back metrics of a set of callers of a set of historical calls, a set of agent records of a set of agents of the call center that participated in the set of historical calls, and time series forecasting, wherein training the predictive machine learning model includes applying regularization for a plurality of iterations;
retrieve data associated with a caller of an inbound call of a call center between the caller and an agent of the call center, the data comprising at least an interaction of the caller with an interactive voice response unit (IVR) during the inbound call and an agent record associated with the agent of the call center;
in response to determining that the inbound call has terminated:
execute the predictive machine learning model, using the data including the interaction with the IVR during the inbound call and the agent record of the agent, to generate output data comprising (i) a call-back signal corresponding to a likelihood of repeated call-backs by the caller, and (ii) an interaction signal representative of a likelihood that the caller will perform a predetermined interaction during a future call-back, as a result of the termination of the inbound call between the caller and the agent;
classify, by the predictive machine learning model executing based on the call-back signal determined by the predictive machine learning model, the caller into one of a first call-back group and a second call-back group;
in a first event where the predictive machine learning model classifies the caller into the first call-back group, instruct the automatic calling device in communication with the processor to automatically initiate an electronic communication session between the caller and at least one agent of the call center; and
in a second event where the predictive machine learning model classifies the caller into the second call-back group, instruct the automatic calling device in communication with the processor to automatically execute a subordinate call-back procedure.
9 . The system of claim 8 , wherein the inbound call is terminated based on the caller exercising a call-back option or by abandoning the inbound call.
10 . The system of claim 8 , wherein the data associated with the caller comprises at least one of an interaction of the caller with at least one advisor of an enterprise associated with the call center or call-back data associated with the caller.
11 . The system of claim 8 , wherein the electronic communication session is initiated between the caller and a preferred agent of the call center.
12 . The system of claim 8 , wherein the call-back signal further corresponds to a likelihood that the caller will accept an offer to purchase one of a plurality of products.
13 . The system of claim 8 , wherein the call-back signal further corresponds to a likelihood that the caller will lapse in payments for a purchased product.
14 . The system of claim 8 , wherein the subordinate call-back procedure comprises one or more of automatically calling back with a recorded message of the IVR, automatically calling back for connection to an agent from a pool of unskilled agents, and automatically deferring call-back.
15 . A system comprising:
a computer-readable medium having a set of instructions that when executed cause a processor to:
train a predictive machine learning model using call-back metrics of a set of callers of a set of historical calls, a set of agent records of a set of agents of a call center that participated in the set of historical calls, and time series forecasting, wherein training the predictive machine learning model includes applying regularization for a plurality of iterations;
retrieve data associated with a caller of an inbound call of the call center between the caller and an agent of the call center, the data comprising at least an interaction of the caller with an interactive voice response unit (IVR) during the inbound call and an agent record associated with the agent of the call center;
in response to determining that the inbound call has terminated:
execute the predictive machine learning model, using the data including the interaction with the IVR during the inbound call and the agent record of the agent, to generate output data comprising (i) a call-back signal corresponding to a likelihood of repeated call-backs by the caller, and (ii) an interaction signal representative of a likelihood that the caller will perform a predetermined interaction during a future call-back, as a result of the termination of the inbound call between the caller and the agent;
classify, by the predictive machine learning model executing based on the call-back signal determined by the predictive machine learning model, the caller into one of a first call-back group and a second call-back group;
in a first event where the predictive machine learning model classifies the caller into the first call-back group, instruct an automatic calling device in communication with the processor to automatically initiate an electronic communication session between the caller and at least one agent of the call center; and
in a second event where the predictive machine learning model classifies the caller into the second call-back group, instruct the automatic calling device in communication with the processor to automatically execute a subordinate call-back procedure.
16 . The system of claim 15 , wherein the inbound call is terminated based on the caller exercising a call-back option or by abandoning the inbound call.
17 . The system of claim 15 , wherein the data associated with the caller comprises at least one of an interaction of the caller with at least one advisor of an enterprise associated with the call center or call-back data associated with the caller.