The complexity of real world problems today are pushing the limits of software architectures. In this paper we present a novel software architecture that enables aggregation of data from multiple sources while giving granular access and privacy to data via access policies. The proposed software architecture is evaluated in two real life use cases. The first use case creates a large data set of scientific articles by aggregating three data sources. Citation lists of scientific papers are analyzed in order to quantify and understand the impact of the work, tracing the ”footprints” of the author in science. The second use case analyze an Automatic Identification System (AIS) large data set, proposing a data reduction technique specific to AIS data. Granular access is provided to third party consumers. Through these we see that the proposed approach is able to easily enriching data knowledge and can provide granular data access to third party consumers.