In the past decade cloud computing has become an essential tool in the storage and analysis of data. Examples include weather prediction, content delivery, and image processing. Conse- quently, security and confidentiality of sensitive data in the cloud has become an important topic in the cloud computing research space. Previously Tetali et al. presented MrCrypt, a secure single cloud framework that transformed cloud based MapReduce programs using partially ho- momorphic encryption. Dong et al. extended this research by using novel data analysis, as well as a trusted client to improve the efficiency of the overall framework. The work presented in this paper aims to extend the research of Dong et. al. by further parallelizing the single trusted client and use two different clouds to preserve data confidentiality.