This paper presents the structure and functionality of zFunction, which is an adaptive distributed computing platform that supports a user-friendly programming model for developing parallel processing applications. It allows developers to design software as if they are programming for a single computer and then it automatically takes care of data distribution and task parallelization activities on different cluster nodes or multiple CPU cores. zFunction thus substantially improves the performance of complex distributed applications that process a large amount of data in real time, mission critical systems. This paper uses a representative case study from the financial services domain to show how these types of applications can benefit from zFunction.