We propose a new, streamlined, two-step geographic visual analytics (GVA) workflow for efficient data storage and access based on a native web XML database called TreeTank coupled with a Scalable Vector Graphics (SVG) graphical user interface for visualization. This new storage framework promises better scalability with rapidly growing datasets available on the Internet, while also reducing data access and updating delays for collaborative GVA environments. Both improve interactivity and flexibility from an end-user perspective. The proposed framework relies on a REST-based web interface providing scalable and spatio-temporal read-write access to complex spatio-temporal datasets of structured, semi-structured, or unstructured data. The clean separation of client and server at the HTTP web layer assures backward compatibility and better extensibility. We discuss the proposed framework and apply it on a prototype implementation employing world debt data. The excellent compression ratio of SVG as well as its fast delivery to end users are encourageing and suggest important steps have been made towards dynamic, highly interactive, and collaborative geovisual analytics environments.