| CPC H04M 3/4936 (2013.01) [H04M 3/42221 (2013.01); H04M 2201/40 (2013.01)] | 18 Claims |

|
1. An interactive voice response system, supplemented by a complaint differentiation and sentiment model, the interactive voice response system comprising:
a transceiver unit operable to receive and respond to audio calls;
an automatic speech recognition system (“ASR”) system operable to:
receive audio calls from the transceiver unit;
transcribe the received audio calls; and
transmit the transcribed audio calls to the complaint differentiation and sentiment model;
the complaint differentiation and sentiment model operable to:
analyze the transcribed audio calls for complaints;
determine that the transcribed audio calls indicated a complaint;
process the transcribed audio calls and identify a first plurality of call components, said first plurality of call components comprising complaint metadata and call content;
transmit the transcribed audio calls that indicated a complaint and the first plurality of call components to an insight dashboard;
the insight dashboard operable to:
display data relating to the transcribed audio calls;
aggregate the first plurality of call components to a second plurality of call components already-stored at the insight dashboard, said second plurality of call components comprising complaint metadata and call content;
display the second plurality of call components;
based on the pluralities of call components, create and display a plurality of graphs, said plurality of graphs comprising:
a complaint graph operable to display a trend in registered complaints per unit time;
a resolution graph operable to display a trend in resolved audio calls per unit time;
an escalation graph operable to display a trend in escalated audio calls per unit time;
a fraud graph operable to display a trend in fraudulent audio calls per unit time,
a statistical graph operable to display a comparison between the registered complaints per unit time, the resolved audio calls per unit time, the escalated audio calls per unit time and the fraudulent calls per unit time; and
a comparison graph operable to compare a total number of audio calls to resolved calls and escalated calls;
using the plurality of graphs to identify one or more audio calls associated with an unresolved complaint; and
the complaint differentiation and sentiment model is further operable to schedule a follow-up communication to a user, associated with the one or more audio calls associated with the unresolved complaint.
|