Do you know that our enterprise mash up point&click solution can be powered by embedded or distributed in-memory data fabric store for analytic purposes?

This is relevant because up-to-know, usually most of the dash-boarding  software for data visualization rely on the infrastructure (hardware, underline database and appserver) for running CPU/memory intensive complex analytic. However, one can add 10Tb in-memory store (one or n nodes) for running analytic purposes, and build a mash-up after it.

Have a look at the following code:

<variable name="congress" type="document"/> 
   <loadfrom cache='sample-cache.LegislatorsDeclCache' 
   fetchsize='10000' variable='congress' />  
  <raql outputvariable='result'> 
    select firstname, lastname, state, party, gender from congress 
    order by state, gender 
  </raql> 
</mashup>

Previously we have defined, for example,  a “BigMemory” data source to store all the USA legislators activity,  bind this big data source to a local variable, and then produce a “result” HTML5 output variable using Raql language. Then one link this output to any type of display diagram, and do the filtering, order by, group-by, etc.

Really, really powerful and extensible!

More info for business users here

http://www.softwareag.com/corporate/products/bigdata/apama_analytics/products/presto/overview/default.asp

and detailed info for techs users here

http://mdc.jackbe.com/prestodocs/v3.7/raql/raqlIntro.html

Here you have the JackBe Presto architecture powered by Terracotta BigMemory add-on:

Presto Analytics Components
Presto Analytics Components