In a typical data factory environment, the data processing subsystem looks the most like a traditional data warehouse. This is where data from all sources get ground into a single version of the truth. It generally consists of a staging database where raw data extracts and data acquisition files are stored. The database is then transformed and loaded into an appropriately designed data model. ETL platforms, as their name (Extract, Transform, Load) suggests, provide a number of tools that automate this process and the better ones also include facilities for lifecycle management, version control and error checking. But even with these ETL tools, there is a fair amount of manual scripting that needs to be performed and maintained.