The big challenge of today in parallel and distributed computing is how to deliver the computing as a service and how to enable the processing of the large-scale data. Especially promising to address both challenges is the combination of cloud computing and big data analytics. The cloud computing technology aims at the provision of compute resources as a utility, the big data analytic is a term addressing methods of data intensive, massively parallel processing in distributed, shared nothing architectures. The marriage of those two technologies is a logical consequence of their technological compatibility and complementary objectives. In truth, the cloud infrastructure provides a user with capability to collect, store and organize their data, while the big data analytics helps him to make sense of all the collected data, while still profiting from the elasticity of the cloud, the possibility to expand or shrink the resources as needed. The proposed project addresses a set of fundamental research challenges posed by the concepts of the cloud computing and the big data analytics and their symbiosis. We target the problem of resource management in cloud infrastructure, we propose to apply the mobile, autonomous agent methodology to deal with its complexity; we address the interoperability issues in Cloud by the means of the resource abstraction. In the domain of the big data analytics, we concentrate on the challenging problem of large scale, distributed graph data management and its integration with other big data analysis approaches. The graph data processing is especially challenging due to its random access pattern and the data driven computation model. We propose to design a distributed graph database system that will enable us to tackle the large scale graph processing and exploit it to address fundamental issues in text processing entwined with the network data.