The Industry Analytics Kit allows consists of a set of re-usable and scalable Analytics that perform a range of streaming analytics over event streams. The scope of the Industry Analytics Kit is that they consume and generate defined "Data" events. Integration into and out of a Analytics network is not defined and allows for integration with any system.

A Data event is a normalized event structure which will be used within the framework for passing around values. Events coming in from external sources will need to be converted prior to use within the framework.

The intent is to have a set of delivered computation Analytics which can be chained together as desired with the framework, and to allow users to easily create their own Analytics. It is anticipated that the modules will largely be EPL monitors rather than EPL queries largely due to performance reasons, especially with regard to memory use and the need to potentially keep large amounts of data within potentially long windows. It seems likely that using techniques such as EWMA (Exponentially Weighted Moving Average) will be more than sufficient for the analyses needed whilst being far more efficient in terms of the amount of individual data points which need to be tracked at any given point.

Analytics will have to have the ability to listen for Data events, so that they can use incoming raw data, and the ability to send Data events so that EPL queries and other components can be used if required.