7 Steps to a Data-Centric Nirvana

Let’s be honest: Data has not been nearly as much of a transformative force as we thought it would be. Four decades into the digital era, we have more data generation and dissemination capabilities than ever, we generate and gather more data than ever, and yet, many basic business practices haven’t changed since the ’80s. Example: Was it really so long ago that you made a copy of a document just to get work done?

Here's the problem: Our data foundation is ultimately application centric. There’s an app for everything, and a database for every app. The average enterprise uses dozens of applications, creating hundreds of silos, undergoing thousands of integrations.

There’s always a relentless pursuit of new technologies delivering new capabilities to meet needs that didn’t exist before but keep emerging. That’s good, but it isn’t the data-enabled transformation we dreamed of -- it’s a technology quagmire that’s always getting more fragmented, complex, risky, and costly.

The only way to radically change this dynamic is with technology and strategy -- we need to separate the data from the applications. This is how we enable real data collaboration that doesn’t undermine security or compliance (definitely no more copying), essentially make data integration obsolete (it can drain half the IT budget) and speed the development of new solutions. Put simply, we need to create a technological world that revolves around data access, not copies.

Related:Keeping Up With Data Privacy Compliance: A Guide

There’s a new generation of technologies that decouples data from the applications used to create, store, and protect it. With these capabilities, we can take a different direction that makes data -- not those apps -- fundamental to the operation. This is real data-centricity, and also real data virtualization.

I was blown away when I saw an actual data collaboration platform from Cinchy in action. It’s an intuitive construct for the way data can be centrally managed, accessed, used, created, and updated. It simultaneously transforms and simplifies application development, and eliminates the need for data copies, silos, and integrations.

This model of data collaboration will change the way we think about development and increase the speed and velocity with which we deliver. It will also dramatically accelerate artificial intelligence and machine learning projects, increasing their accuracy while allowing AI to scale, with full control over the data. The possibilities truly are mind-boggling, and endless.

Moving from app-centric to data-centric/Image by Sylvie Veilleux

To get to that data-centric nirvana, here’s an outline of the steps we might take along the way.

1. Manage data as a network: This redefines plug and play. All forms of data -- HR, finance, marketing, product development and legal -- becomes interconnected, just like we expect our devices to be. In a world of hardware and software, this gives data top billing, and ends point-to-point data integration.

2. Get to zero-copy status: Now an emerging standard, Zero-Copy Integration is already gaining traction in Canada and elsewhere. This enables all data to be shared via a single, central collaboration platform. Goodbye data copies, duplication, endless IT integration projects, costs, and complexity; hello, business enablement, time to market, improved risk and compliance posture.

3. Set data free (as in, ‘liberated and autonomous’): This lets data exist outside of any systems or applications. When data is decoupled from applications, it enables better control and governance, aids in the creation of metadata (data about data), and finally allows for controlled co-production of data. Data products, in a meshed environment where control is at the data level and where data can be both viewed and edited, directly as needed.

4. Achieve collaborative intelligence: When data is autonomous and accessible across projects, systems, and people, a technology platform becomes an ideal environment for collective wisdom and collaborative intelligence. In the case of a merger between two companies with disparate systems, it would mitigate a plethora of headaches.

5. Strive for schema plasticity: For all the advances we see in technology, many painful issues haven’t changed at all—just think of the proliferation of spreadsheets. And when a developer changes the underlying schema or framework of an application, it can disrupt every kind of functionality. This may require changing the operating model entirely. Decoupling data from applications rids developers of these issues and gives them the time to innovate–we can bid farewell to many forms of regression testing or analysis, and measurably speed the development of, say, custom applications needed to take advantage of business opportunities.

6. Enrich the experience with metadata: We can think of applications, websites and social channels as skins covering the data, and we already know they’re all largely intertwined. But as we separate the information from the wrapping, we’re able to use metadata to create deeper and more valuable experiences. This is another revolution in the works: It gives us the foundation to create richer experiences than ever before in a fraction of the time, beginning with data management and encompassing data discovery, research, analysis, customer experience and insights, data control and governance.

7. Make data the boss: Yes, data is the network, but it’s also the application, and the data product. It’s the boss. We’ve accepted the idea that data serves applications; let’s instead say applications serve the data. Treating data in a way that recognizes its primacy is the only way to turbo-charge IT, development, and just about all business solutions.

Again, it all starts with separating data from applications. This removes the barriers we’ve so methodically erected, which continue to constrain us.