With many different types of organizations depending on geodatabases, the options for architecture are also quite varied. While some organizations have moved all of their data to the cloud, storing data and analysis tools on different servers, most maintain an on-premise geodatabase as the enterprise system. 

A third architecture style, which balances between cloud-based and local geodatabases, is also very popular. This allows for database backups to be supported by always available cloud services and for data services to reach outside of organizational firewalls while limiting the datasets and services that are exposed to the internet.

The balance between these solutions depends on the need for processing speed and storage costs. MapD, which can be installed and maintained locally or can be hosted in the cloud, fits all kinds of organizational requirements. The speed of queries and data processing allows cloud data resources to be used in the same manner as locally-stored datasets. With pymapd, datasets can easily be mirrored in the cloud while maintained locally and can be integrated into geospatial analyses by comparing locally stored data to cloud-based data. 

The technological structure your organization chooses will depend on your needs and the size of the datasets both produced and ingested from other sources. MapD can become a part of this structure or can be the entire GIS, supporting Spatial SQL queries at blazing speeds whether located on-premise, in the cloud or both.