In recent years, almost every industry especially digital, e-commerce and telecommunication experienced exponential growth of data. Harvesting knowledge in these highly dynamic databases and finding closed patterns to analyze modern trends attracted considerable interest in this decade. To group similar types of data object or identify regions where data density is high in a dataset and forming clusters is the birds eye of researchers. Data mining and Business analytics have became an integral part of Telecommunication industry for extracting usage patterns of new as well as profiled customers and retaining them in competitive market. Providing better services without interruption is an essential aspect as this gains the confidence of the customers. In this paper, we have stated various challenges faced by the industry and have proposed a unified architecture for Telecom data analytics. Case studies have been put forward for Customer usage pattern detection and Network fault analysis using clustering and bi-clustering techniques to group and segment customer for business predictions as well as to identify faulty nodes in the network and predict network failures.