In an organization, data is stored in various disparate systems. Data within these systems is referred to as internal data of the organization. Also, with digitization of the world, a lot of external data is available and accessible. Combining internal data with external data can bring insights that have the potential to build new business models or substantially improve existing business models, for companies to stay ahead of competition.

The primary objective of an enterprise data architecture is to make this process easy. Enterprise data architecture is a set of strategies, blueprints and frameworks that enable businesses to quickly onboard and accelerate their journey to become data driven.

The process of building an enterprise data architecture involves, on a high level, the following steps.

Defining Data Architecture Principles

A set of guiding principles enable organizations to build data architecture across the enterprise in a consistent manner. These include guiding principles for data acquisitions, data storage, data cataloging, data organization, data access, data reporting and visualizations, data security, analytical models etc. Guiding principles are also supported with technology frameworks, methodologies, and design patterns.

Analysing AS-IS state of Data and Systems

It is impossible to build an enterprise data architecture, if we are not aware of the existing landscape of systems that hold and manage data within the organization. Analysis and documentation of the existing systems is necessary, which includes identifying key stakeholders of existing systems, their applications, broad definition of data within these systems, constraints and limitations etc

Defining TO-BE Data Architecture

A To-be data architecture has to be aligned with the overall enterprise data management strategy, and should detail out stakeholders, systems, processes and applications. It should have flavours specific to various stakeholders that are involved in building the new architecture – like a business view, a functional view, a systems view and an infrastructure view.

Defining Data Architecture Roadmap

The to-be architecture can be ambitious, and can span across several years. It has elements that are urgent, and elements that are important in the future. A roadmap details out a plan to build various elements within the to-be architecture across time, carefully considering various priorities, dependencies and cost constraints.

An enterprise data architecture should begin with building strategic architecture blueprints, technology competencies and technology frameworks keeping the strategic architecture in mind.

Defining Data Architecture Governance Strategy

Once we start the process of onboarding individual projects according to the defined priority, data architectures should be audited and any architectural exceptions should be considered as architectural debt, and should be monitored and tracked on an ongoing basis.