With ever increasing volume of contact centers globally, there is a massive flow of customer service interactions into contact center telephony systems. These interactions are a rich source of information that can be utilized to uncover valuable insights regarding customer preferences and pain-points. In this paper, we propose an unsupervised pipeline comprising of an ASR engine, a reason-utterance detector, a re-writing module, a topic modelling block and a topic-description generator to enable Voice of Customer (VoC) Discovery by systematically tracking the reasons for the conversations. The insights surfaced by our pipeline can help contact centers get a bird's eye view of what is happening in their business. This would not only help business leaders come up with strategic measures to provide high quality and quicker resolutions to customers but also drive better customer satisfaction.