Application Launcher

Area	AgilOne JIRA	Description of functionality after the fix
Platform	AGO-1169	We squashed a bug that sometimes resulted in duplicate master customer records being generated by the Likelihood to Buy machine learning model.
360 Profiles	AGO-11731	In order to improve 360 performance (only minorly) and reduce response complexity, we have removed all ‘layout’ metadata from the 360 API response. This metadata previously was utilized by the 360 Profiles user interface, but is now stored elsewhere.
Actions	AGO-11817	We squashed a bug in the ‘Purchased a Number of Products’ Actions filter that was deflating the number of products purchased. Previously, this filter could incorrectly de-duplicate transaction items that were in fact unique and should not be de-duplicated. This led to fewer customers being included in the audience since some purchased quantity was not included in the sum of purchased products.
Platform	AGO-11614	If you use Google Ads for digital marketing, you can now use AgilOne to send offline transactions (i.e. conversions) and events to Google Ads for Google to attribute to your digital marketing efforts. This will provide you a more thorough picture of how often Google Ad campaigns influence offline (i.e. physical or retail) transactions.
Actions	AGO-11734	We have introduced a new feature that eases the development of custom Actions audience filters and content models. There is still significant work to do here, but we are making incremental progress.
Platform	AGO-11769	We have improved the performance of our nightly Identity Resolution Engine (IRE) that de-duplicates source customer records into master customer profiles. This should quicken the nightly batch refresh process as several downstream steps are dependent upon IRE.

1904 / 1906 Release - 2019-02-25