Advanced Computing in the Age of AI | Tuesday, March 19, 2024

The API Economy is Begging for Digital Interoperability: Enter Data-Centric Integration 

“Transformation” is on everyone’s lips – but what does this really mean? In the past, CIOs were preoccupied with aligning IT with the business, but now there’s a sense of urgency about embedding IT into the business, improving the experience of end customers (including internal customers) and driving real business value.

This all starts with understanding how application integration, process management and data integration – once quite separate disciplines, with siloed technology – are converging. This convergence is necessary to derive value from the hybrid architecture of today’s modern enterprise. Here is what each of these means:

Data Integration
It has long been said that data is the new oil, and now organizations are spending significant sums finding value within their vast data lakes. Data integration has typically played a role at this layer, combining data residing in different sources and providing developers and users with a unified and/or simplified view of this data. This is a key function that most integration platforms (iPaaS) simply ignore.

Process Integration
As organizations embark on a journey of transformation, it is often business processes that teams and leaders take aim at. Legacy processes often slow down transformation and innovation. Predictably, there have been many attempts to simplify and improve these processes - with tools like Business Process Management or other proprietary integrated development environments – but the track record of such tools has been mixed. The latest round of low-code/no-code tools for integration and business process are designed to put integration in the hands of more users, not just expert integrators.

Application Integration
Historically, applications were the primary structural component within the enterprise. Most applications of the 1980s exist today as large monoliths that needed limited integration. Through the late 1990s and 2000s, all manner of middleware was introduced to solve application integration. But organizations are still stuck with these same patterns today – even as the number of applications swells and the complexity of business operations grows exponentially,

With each of these a separate discipline, major cracks have appeared in today’s enterprise architecture. The problem, however, stems the focus being placed on the wrong structural element – interfaces, rather than data.

Enterprise integration architecture is characterized by a portfolio of application systems with clear ownership of their domain. Can you say the same things about application data? For example, are there clear owners for customer data or product data? Are you routinely monitoring and improving the quality of enterprise-shared data? Are you prepared to handle the exponential growth of machine data from new internet-connected devices?

Anyone can buy an ERP – but no one else has data showing what your customers purchased from you, the results of your internal product quality tests, or financial data about your performance. It’s the data, and not the applications, that provides competitive advantage.

Data-Centric Integration
Forward-thinking businesses must embrace the convergence of application, process and data integration, ensuring consistency and governance across hybrid environments. If businesses want the right kind of data to underpin advanced business processes or to create multi-dimensional views of data objects, data-centric integration must be pursued as a strategic function that aligns with business objectives.

Unfortunately, many enterprises today have become bogged down with legacy iPaaS platforms. With enterprises today having on average 1,071 cloud services in use around the world, these legacy integration patterns simply can’t scale. Data-centric integration addresses this problem but turning the focus of application integration towards the data upon which organizations rely, rather than “point-to-point” integration patterns that dominate the integration landscape today.

A New Enterprise Data Model

The first step with data-centric integration is to define the data you care about. But this isn’t the legacy approach to “master data management,” where data models are inflexible and defined by IT rather than business owners. This new enterprise data model is about understanding where value exists, what needs to be shared and what needs to be extended with partners and customers.

Start by thinking about the value your business delivers to your customers: what are the things that make your business work and make it great? Answering these questions often inform the data constructs you’ll want to model. Note that this model isn’t describing data entities, cardinality and other characteristics. Instead, this new model reflects how business functions create and use data in the context of day-to-day operations.

Today, the iPaaS market uses a point-to-point integration patterns that don’t scale. Since 2006, there was a 758% increase in public web APIs. Data-centric integrations – versus interface-focused – enables IT and developers to build connections faster, with data at their core.

Transformation with Data at the Core

Data is strategic; both data that you own and data that you don’t. The ability to bring the right data together, at the right time can enable a business to make decisions around almost perfect data. One of the biggest barriers to effective cloud adoption is integrating, synchronizing and relating data, applications and business processes among cloud and on-premises systems. The added complexity of hybrid makes it difficult to effectively share and use business data across the organization.

Legacy iPaaS platforms alone are no longer sufficient to tackle enterprise cloud integration. Instead, a data-centric approach allows you to define, model and manage the data you care about, rather than the interfaces and systems that produce and consume this data.

By Mark Geene, CEO and Co-Founder, Cloud Elements.

EnterpriseAI