Before I go into why you may need to pay close attention to this new way of generating advanced analytics applications, it’s important to lay the foundation with a brief explanation of what the App-On-Demand premise actually is…
App On-Demand is a method through which a brand new analysis is generated based on ad-hoc user actions within an app. It’s a way to carve out a subsection of an analysis, run advanced analytics, pre-compile as needed, combine with any form of outside data feeds, and reassemble into a new analytics application all without needing to know how any of that other fun stuff may actually work.
Take, for example, the following use case:
Say you are in charge of the group responsible for finding new locations for your nation-wide financial institution. As part of your role you need to identify key markets where your services would thrive, and areas where they may need to be downsized or removed. You do have a rather robust set of analysis tools at your disposal, but there are many factors left unknown. You begin a market analysis based on an area where you want to expand, select the ideal demographics which you know is your target customer, and click on “Run Analysis”
You are shown a map with points, regions, boundaries, estimates and details about the local population and their known affluence, credit health, and levels of home ownership. These points are placed in ideal settings because the analysis that was run in the background also identified the fact that your branches do much better when situated next door or close by a coffee shop and at least a certain distance from your competition. Further enhancing this discovery is that crime rates are decreasing, employment is on the rise, and new construction permits have been increasing at a steady rate for over 2 years. Now, armed with all this data in your on-demand analysis, you can view estimated business for any of the services you provide and thus make a truly informed decision.
So what just happened?
You were provided the ability to dynamically generate and analyze new data on the fly, by way of user-specified criteria. You were shown precise answers in many cases, and in other cases they are estimates based on predictive analytics and targeting through services like R or Compellon. Terabytes upon Terabytes of data were traversed by calls out to APIs such as Alteryx, then returned and assembled into a new application that was built upon a template.
It’s important to keep in mind that without this technique, it would be nearly impossible to compute and compile all potential permutations of geospatial and outlier possibilities, this is why App On-Demand is so important; even traditional Qlik applications cannot accomplish this, due to volume, complexity, load time, etc.
Bardess has great success delivering this solution for several of our customers with both front-end custom Qlik Sense extensions as well as back-end server integration. We can integrate with the rules based engine to control additional items like security, governance, and proxy conditions that may not be available to the end users, all in an effort to make this solution the most viable and complete set of advanced analysis for your company. Because of our broad partnerships and integration relationships, we can deploy a big picture perspective that ties many services together, from data prep with Paxata to CRM predictions with Compellon, and beyond.