Digital transformation is driving organizations to implement data-driven solutions that transform business outcomes. Organizations tend to operate in silos and the same is true of the data and information assets that a company owns. These silos are created by applications. To break down these silos, a data-centric architecture is required for enterprise resilience where data is the primary asset, while applications, tools and use cases may come and go.
The shift from app-centric to data-centric architecture
The explosion of app-centric enterprise architecture has created a universe where new solutions require customized access controls, lengthy integrations and heaps of data duplication. Development of every new solution or capability calls for a forced ramp-up task of copying data and integrating systems. In addition, every application has its own model for data, and developers have had to build around that model adding development time and complexity.
As a result, enterprise architectures are inconsistent, and enterprise IT teams do not have the liberty to make changes to legacy systems, fearing they might break something critical. Agile and forward-looking businesses cannot operate under such constraints as applications are valueless without access to and an understanding of data these applications handle.
Andy Packham, Senior Vice President and Chief Architect in the Microsoft Ecosystem Unit at HCLTech comments: “Data is the oil driving the new economy. To take advantage of the enormous amount of data many enterprises have collected, these enterprises are seeing value in looking beyond a simple application view [where data is accessed through the application] to rapidly re-tool data and view it as a separate asset. Artificial intelligence and machine learning technologies are helping businesses extract value and insight to make the most of their data.”
Elaborating on the concept of data centricity, Packham adds: “Many of us now own a wearable device that collects considerable physiological information–pulse, HRV Heart Rate Variability), movement, blood O2 saturation, breathing rate and more. At a personal level, this information is useful and accessed through an application, usually on a smartphone or other handheld device. However, during COVID researchers could show that, by taking insights from the combined data from many sources, they could predict who may be infected with COVID before symptoms developed—and prompt the individual to get tested. Without creating insights from multiple sources of data, this wouldn’t have been possible.”
Operationalizing data-centricity with artificial intelligence
According to Harvard Business Review research of over 1500 companies, firms achieve the most significant performance improvements when humans and machines work together. Through collaborative intelligence, humans and AI actively enhance each other’s complementary strengths: the leadership, teamwork, creativity and social skills of the former, and the speed, scalability and quantitative capabilities of the latter. However, achieving this synergy is only possible through data-centricity along with a data collaboration platform rather than application-centric platform.
Data-centricity is a mindset as much as it is a technical architecture — at its core, data-centricity acknowledges data’s valuable and versatile role in the larger enterprise and industry ecosystem and treats information as the core asset to enterprise architectures. Opposite of the “Application-Centric” stack, a data-centric architecture is one where data exists independently of a singular application and can empower a broad range of information stakeholders.
Data-centricity democratizes data, giving data owners unprecedented control and enabling new solutions and business insights. It removes the barriers that make AI difficult to work with, paving the way for AI-powered solutions to revolutionize the way businesses work.
“Migrating to a data-centric approach isn’t just a single decision, it is a product of many decisions made by leaders and is as much about changing culture and business models as it is about the technology selection. A data-centric mindset shifts the focus on understanding and improving data, rather than the model architecture. So, to improve the performance data needs to be understood and labeled and needs to be complete and representative to avoid minimizing data bias,” adds Packham.
Real-world use cases require subject matter expertise along with private data and rapidly changing objectives. Many ethical and governance challenges are aggravated by AI approaches relying on manual labeling and to overcome this challenge, enterprises need to know:
- How to inspect, obliterate and minimize biases
- How to collect representative data keeping in mind fairness and inclusivity
- How to govern or audit labeled data points
- How to trace the lineage of model errors originating from incorrect labeling
The security challenge in cloud migration
Focusing on data changes the nature of cybersecurity challenges.
“Cyber-attacks and data breaches are one big challenge. Organizations need a better way to protect their sensitive data internally. Legacy security technologies focus on the location of the data (endpoint, server or network), whereas data-centric security identifies sensitive data and applies policy-based protection to secure that information throughout the data lifecycle, irrespective of its location,” says Packham.
He continues: “Insights from data come from combining many sources, some of which may be critical to the success of the business but might be created and owned by an entirely different organization. So, as with challenges in the physical supply chain, enterprises should consider the risks and controls needed to establish a successful digital supply chain.”
HCLTech partnering with Microsoft Azure to deliver data-centric applications
With over 30 years of partnership with Microsoft and 8,500 certified professionals on Azure, HCLTech has been helping companies achieve a meaningful transformation in the Microsoft Cloud.
Packham explains: “Azure makes data-centric transition in the cloud easy, and that can be part of the problem. Data-centricity is not about collecting and storing massive amounts of data. There are costs and risks involved in managing that state. However, with the right partner who understands not only the technical landscape but the business ecosystem these challenges can be avoided.
"Moving to data-centricity is complex, and leaders need to recognize that this is a journey, not an event. Organizations must think holistically about the business outcomes, ethics, compliance and security.”
Microsoft Azure provides a complete solution, from the infrastructure layer that imposes controls to the Intelligent Data Platform that provides a scalable and secure storage, data governance and AI/ML-driven insights platform. But the strength is the breadth of the entire Microsoft Cloud. Developers can take advantage of the Power Platform to build their own low-code/no-code applications that can deliver information directly to teams.
Packham adds: “Very broad capabilities are needed to create a successful data-centric platform and help create true business value while maintaining security and ensuring ethical and transparent usage of information. HCLTech is one of the first Microsoft partners to have been awarded all six Solution Partner designations. This recognizes the depth and breadth that HCLTech has across Infrastructure, Data & AI, Digital & Application Innovation, Modern Work, Business Applications and Security. You can’t only focus on one area and be truly successful.”