Rating-based collaborative filtering (CF) enables the prediction of the rating that a user will give to an item, based on the ratings of other items given by other users. However, doing this while preserving the privacy of rating data from individual users is a significant challenge. Several privacy preserving schemes have, so far been proposed in prior work. However, while these schemes are theoretically feasible, there are many practical implementation difficulties on real world public cloud computing platforms. In this paper, we approach the generalised problem of privacy preserving collaborative filtering from the cloud perspective and propose an efficient and secure approach that is built for the cloud. We present our implementation experiences and experimental results based on the Google App Engine for Java (GAE/J) cloud platform.