A common question is the difference (or similarity) between Data Warehouse Automation and traditional ETL tools.
I like to use an example from my iPad – the difference between the apps Mortgage Calc andPages.
Pages is a spreadsheet. You can edit rows and columns of data, and create formulae using that data.
Mortgage Calc is an app that calculates mortgage payments.
Now, I could write a mortgage calculator in Pages. I could possibly make it look like the Mortgage Calc app rather than a spreadsheet. But which calculations do I use? Which tax rules apply? Are there stamp duties payable? In other words, I have to do a lot of research, a lot of programming, and a lot of testing to make sure I’ve got the basics right. And I’ve also got to maintain that spreadsheet as the rules change.
With Mortgage Calc, I’ve paid a few dollars for an application that has saved me many hours of research and development, and which I’m trusting to give me accurate calculations. In this case, Mortgage Calc is better than Pages, because it does one job, and does it well.
That is the difference between DWA and ETL. An ETL tool is a general purpose programming environment for moving and transforming data between systems. It provides components, in one form or another, which you put together to accomplish one or more tasks.
DWA, on the other hand, is built to do just one task, which is building the code associated with a data warehouse. Ajilius builds dimensional data warehouses. We build transactional, periodic snapshot and accumulating snapshot fact tables; Type 0, 1, 2 and 3 slowly changing dimensions; and move data from multiple data sources into a consolidated presentation layer.
You could do all of that with an ETL tool, but it would be like writing a complex mortgage calculator in a spreadsheet – time consuming, not well understood, and prone to error.
Ajilius generates fast, error free code, that can be easily migrated between data warehouse platforms, at the press of a button.
That’s the advantage of tools like Ajilius. We deliver business value, faster.