I’m proud to announce that, today, we achieved our goals for Ajilius 2.3.
The hero feature of this release is data profiling. We now profile data sources faster than Trifacta, with more valuable information than Pentaho, and with more variety than SSIS. We give you the real data you need to make quality decisions about the content of your data sources.
To deliver this feature, we first added persistent metadata caching to Ajilius, as discussed in this earlier post.
Now, we’ve completed the feature by implementing the profiling and presentation features.
You profile a source table from the Extract Source Tables screen. The following screen shot shows that we are about to profile the Chinook Customer table.
We see any previous profile that is cached for this table, and we can refresh the profile at any time by pressing the Profile button.
We profile any number of rows, at a rate of around 4million rows per minute.
Every column is profiled in every row.
For columns of less than 64 characters in length, we profile up to 1,000,000 discrete values per column.
For columns of up to 256 characters in length, we profile up to 1,000,000 discrete patterns per column.
For columns of up to 4,000 characters in length, we profile the minimum and maximum values in a column.
Not only do we profile values and patterns, we examine your data for characters that might cause problems in your data warehouse. Null values? Got it. Control characters? Check. Extended ASCII characters? That too. Unicode characters? Again, check.
This is real, valuable profiling data for data warehouse professionals. And it is now included in your Ajilius licence.
So, once again, Ajilius provides real value through the addition of the features you need.
Ajilius. The real innovators in data warehouse automation.