Application Intelligence – Why Should You Care?

5 min read
May 24, 2022 10:01:00 AM

The software industry is world class at introducing new terms that promise to transform how people leverage technology for good in a way that has never been possible before.

In many cases these terms are old ideas, repackaged with a slight twist with a catchy name or a 3-letter acronym. Few of them ever become part of our everyday vocabulary, especially when targeted towards a non-technical audience.

If you struggle to pull data from multiple systems and blend it together then Application Intelligence might become one of those terms worth putting on a watch list.

If you believe in the concept of data democracy, where people’s access to data is no longer limited to their technical capability and where access to data is no longer used to “get one up” on colleagues, you might become an evangelist for Application Intelligence.

Simple term, powerful concept.

Application Intelligence is a simple term for a powerful concept. It looks to simplify how people connect to and blend data from the enterprise applications they use in their business on a daily basis.

Application Intelligence is an additional layer of intelligence that can understand everything about the systems you are looking to pull data from and send data to, quite simply, so you do not have to. Application Intelligence lowers the technical bar for people looking to work with data from multiple systems.

This “built in” knowledge extends beyond how to connect to the systems and includes innately understanding how the data is stored and structured and how it relates to each other. Where possible, Applications Intelligence looks to deal with any undocumented idiosyncrasies or special “features” of the data that only a deep domain expert in that enterprise application would ever hope to understand.

Like most things, should Application Intelligence become successful, other software vendors will look to mimic the approach, but how many will really go to the level required to engineer-out all the complexity and make it so easy that just about anybody can do it?

Time for a mind shift.

If the goal of Application Intelligence is to make blending of data from multiple sources together and streaming the results to your favourite applications so easy that anybody can do it, then there needs to be a mind shift change in how people solve this problem.

  • We need to move away from needing to use a collection of tools and technologies to solve this problem. While technical people enjoy exploring most business users would prefer a single application or platform, a one stop shop.
  • The solution needs to have the breadth of functionality to solve all the problems on the data supply chain. Everything from connectivity to source systems, to transformation of data all the way to streaming the clean data to any application the user wants to work in. There is no point in solving part of the problem if you can't solve all the problem.

The starting point for many users is when they look to connect to a data source or system. APIs (Application Programming Interfaces) have tried to simplify how people can access data that is locked in a system, but this remains a technical activity.

Making something simple is hard.

However, a data blending system built with Application Intelligence makes this a point and click experience. Users can specify what system or source they want to connect to, and it auto-configures the best way to connect.

The user only needs to enter their user details and security credentials and everything else is taken care of. While this sounds easy for a software vendor to build into their software, it requires a lot of work because every system requires different things to establish the optimum data connection.

Once you have established a connection, the first thing a user is faced with is a bunch of instantly unrecognizable field and table names. For many people, this stops them in their tracks, and they go in search of IT help.

Some systems, especially ERP (Enterprise Resource Planning) systems, contain a data dictionary which holds the keys to the kingdom that explain the fields, tables, and their purpose. Any software vendor that is prepared to invest the time to understand the data dictionary can really simplify the puzzle of where a user can find the data they want.

Who knows what you will find.

Enterprise application databases can in many ways resemble a teenager’s bedroom in that they appear disorganized and foreign to the visitor but make perfect sense to the one who lives in it from day to day.

When built, few software engineers ever really expected the outside world would want to access this data outside of the application it was designed for and in turn, never really paid much attention when naming tables and fields.

Application Intelligence leverages Natural Language Processing to transform field names into human-friendly forms. This enables the user to get past what is often the first step and thought - “I don’t know what I am looking at!” This is particularly important given the volumes of that data that need to be blended across multiple systems.

For some reason, people tend to benchmark data preparation, data integration and data blending systems by the number of systems then can connect to. This is fine if you are just moving data broadly untouched from one system to another in a form of data synchronization.

While connections are obviously important when you are looking to blend data from multiple systems, transformation is equally important. Unless you can unify/transform the multiple data sources into a common structure, you are just not going to be able aggregate of blend the data together.

Applying Application Intelligence principles to this challenge means that when you connect to a data source your data blending solution will dynamically alter the application data structure into an instantly usable form. It does this by creating its optimized dataset without touching or impacting the source data.

Unusual date fields can be automatically converted, text turned to numbers where necessary, even decimal shifting and data trimming might also be required and can be done easily. It can also suggest common fields or keys that can be used to link the data.

Typically, these are the types of things only someone with domain expertise would know how to do.

Inbuilt Application Intelligence replaces the need for IT or domain expertise in specific systems. Collectively, this means you do not need to be an expert to make sense of your data.

Behavioral science suggests that people are more likely to try something new or push boundaries if they know they know that they can unwind or fix anything that doesn’t quite do to plan.

Blending data from multiple systems has historically felt above the “technical pay grade” of most end users even the most hardened Excel pivot wizards.

Application Intelligence helps business users make the leap by giving them the ability to see all the changes they have made to the data flow with a clear before and after picture of each change and more importantly, the ability to rewind any change.

It is this digital safety net that encourages users to move beyond what they have previously felt possible and take control of the data they need to do their job.

Only time will tell if Application Intelligence becomes part of your day-to-day vocabulary – but there is little doubt that any data preparation, data integration or data blending solution that understands the idiosyncrasies of the systems they work with offers a significant unfair advantage to those prepared to go further than they have ever gone before in the search of their own data independence.

To see an example of Application Intelligence in action, go to www.eyko.io

Describe your image

No Comments Yet

Let us know what you think