In today's modern era, most of the organizations have their own call centers or they outsourced another company for handling calls. Customers have high expectations that their issues be addressed and handled quickly and efficiently. In call-centers when a customer problem is not resolved by one service person, it is forwarded to the person who is the superior one. Most of the time the call recording of the customer is used for further references. In that case whoever has reviewed the call recording has to go through the entire recording which is a very tedious task. In this paper, we designed a methodology by which we have summarized the audio data. First, we used a Speech recognizer for converting audio data to Text using AssemblyAI API and then summarized it using the Graph-theoretic approach.