The space mission design and operations community already recognizes the value of cloud computing and virtualization. However, natural and valid concerns, like security, privacy, up-time, and vendor lock-in, have prevented a more widespread and expedited adoption into official workflows. In the interest of alleviating these concerns, we propose a series of guidelines for internally deploying a resource-oriented hub of data and algorithms. These guidelines provide a roadmap for implementing an architecture inspired in the cloud computing model: associative, elastic, semantical, interconnected, and adaptive. The architecture can be summarized as exposing data and algorithms as resource-oriented Web services, coordinated via messaging, and running on virtual machines; it is simple, and based on widely adopted standards, protocols, and tools. The architecture may help reduce common sources of complexity intrinsic to data-driven, collaborative interactions and, most importantly, it may provide the means for teams and agencies to evaluate the cloud computing model in their specific context, with minimal infrastructure changes, and before committing to a specific cloud services provider.