What is data architecture? A framework to manage data

Data architecture definition

Data architecture describes the structure of an organization’s logical and physical data assets, and data management resources, according to The Open Group Architecture Framework (TOGAF). It’s an offshoot of enterprise architecture that comprises the models, policies, rules, and standards that govern the collection, storage, arrangement, integration, and use of data in organizations. An organization’s data architecture is the purview of data architects.

Data architecture goals

The goal of data architecture is to translate business needs into data and system requirements, and to manage data and its flow through the enterprise. Many organizations today are looking to modernize their data architecture as a foundation to fully leverage AI and enable digital transformation. Consulting firm McKinsey Digital notes that many organizations fall short of their digital and AI transformation goals due to process complexity rather than technical complexity.

Data architecture principles

According to David Mariani, founder and CTO of semantic layer platform AtScale, six principles form the foundation of modern data architecture:

1. View data as a shared asset. A modern data architecture needs to eliminate departmental data silos and give all stakeholders a complete view of the company: 360 degrees of customer insights and the ability to correlate valuable data signals from all business functions, like manufacturing and logistics.
2. Provide user interfaces for consuming data. Beyond breaking down silos, modern data architectures need to provide interfaces that make it easy for users to consume data using tools fit for their jobs. Data must be able to freely move to and from data warehouses, data lakes, and data marts, and interfaces must make it easy for users to consume that data.
3. Ensure security and access controls. Modern data architectures must be designed for security, and they must support data policies and access controls directly on the raw data, not in a web of downstream data stores and applications.
4. Establish a common vocabulary. Shared data assets, such as product catalogs, fiscal calendar dimensions, and KPI definitions, require a common vocabulary to help avoid disputes during analysis.
5. Curate the data. Invest in core functions that perform data curation such as modeling important relationships, cleansing raw data, and curating key dimensions and measures.
6. Optimize data flows for agility. Limit the times data must be moved to reduce cost, increase data freshness, and optimize enterprise agility.

Data architecture components

A modern data architecture consists of the following components, according to IT consulting firm BMC:

• Data pipelines. A data pipeline is the process in which data is collected, moved, and refined. It includes data collection, refinement, storage, analysis, and delivery.
• Cloud storage. Not all data architectures leverage cloud storage, but many modern data architectures use public, private, or hybrid clouds to provide agility.
• Cloud computing. In addition to using cloud for storage, many modern data architectures make use of cloud computing to analyze and manage data.
• Application programming interfaces. Modern data architectures use APIs to make it easy to expose and share data.
• AI and machine learning models. AI and ML are used to automate systems for tasks such as data collection and labeling. At the same time, modern data architectures can help organizations unlock the ability to leverage AI and ML at scale.
• Data streaming. Data streaming is data flowing continuously from a source to a destination for processing and analysis in real-time or near real-time.
• Container orchestration. A container orchestration system, such as open-source Kubernetes, is often used to automate software deployment, scaling, and management.
• Real-time analytics. The goal of many modern data architectures is to deliver real-time analytics — the ability to perform analytics on new data as it arrives in the environment.

Data architecture vs. data modeling

According to Data Management Book of Knowledge (DMBOK 2), data architecture defines the blueprint for managing data assets as aligning with organizational strategy to establish strategic data requirements and designs to meet those requirements. On the other hand, DMBOK 2 defines data modeling as, “the process of discovering, analyzing, representing, and communicating data requirements in a precise form called the data model.”

While both data architecture and data modeling seek to bridge the gap between business goals and technology, data architecture is about the macro view that seeks to understand and support the relationships between an organization’s functions, technology, and data types. Data modeling takes a more focused view of specific systems or business cases.

Data architecture frameworks

There are several enterprise architecture frameworks that commonly serve as the foundation for building an organization’s data architecture framework.

• DAMA-DMBOK 2. DAMA International’s Data Management Body of Knowledge is a framework specifically for data management. It provides standard definitions for data management functions, deliverables, roles, and other terminology, and presents guiding principles for data management.
• Zachman Framework for Enterprise Architecture. The Zachman Framework is an enterprise ontology created by John Zachman at IBM in the 1980s. The data column of the Zachman Framework comprises multiple layers, including architectural standards important to the business, a semantic model or conceptual/enterprise data model, an enterprise/logical data model, a physical data model, and actual databases.
• The Open Group Architecture Framework. TOGAF is an enterprise architecture methodology that offers a high-level framework for enterprise software development. Phase C of TOGAF covers developing a data architecture and building a data architecture roadmap.

Modern data architecture best practices

Data architecture is a template that governs how data flows, is stored, and accessed across a company. Modern data architectures must be designed to take advantage of technologies such as AI, automation, and internet of things (IoT). According to data platform Acceldata, there are three core principles of data architecture:

1. Scalability. Modern data architectures must be scalable to handle growing data volumes without compromising performance. A scalable data architecture should be able to scale up (adding more resources or processing power to individual machines) and to scale out (adding more machines to distribute the load of the database).
2. Flexibility. Flexible data architectures can integrate new data sources, incorporate new technologies, and evolve with business needs.
3. Data integrity. Modern data architectures must ensure data remains accurate, consistent, and unaltered through its lifecycle to preserve its reliability for analysis and decision-making. They must prevent issues like data corruption, duplication, or loss.

Modern data architectures should also adhere to the following best practices:

• Align with business needs. Effective enterprise data architectures should align with business goals. To do this, organizations should identify the data they need to collect, analyze, and store based on strategic objectives.
• Ensure data governance and compliance. Robust data architectures need to ensure data governance and compliance to establish clear policies for managing data access, quality, and security throughout the data lifecycle.
• Choose the right tools and technologies. It’s critical to select the appropriate tools for your enterprise data architecture, including relational and NoSQL databases, cloud-based storage solutions, and processing tools.

Dan Sutherland, senior director of technology consulting at consulting firm Protiviti, adds more best practices for modern data architectures:

• Cloud-native. Modern data architectures should be designed to support elastic scaling, high availability, end-to-end security for data in motion and data at rest, and cost and performance scalability.
• Scalable data pipelines. To take advantage of emerging technologies, data architectures should support real-time data streaming and micro-batch data bursts.
• Seamless data integration. Data architectures should integrate with legacy applications using standard API interfaces. They should also be optimized to share data across systems, geographies, and organizations.
• Real-time data enablement. Modern data architectures should support the ability to deploy automated and active data validation, classification, management, and governance.
• Be decoupled and extensible. Modern data architectures should be designed to be loosely coupled, enabling services to perform minimal tasks independent of other services.

Data architecture roles

Here are some of the most popular job titles related to data architecture, and the average salary for each position, according to data from Indeed:

• Data architect: $67,000-$173,000
• Project manager: $57,000-$142,000
• Data engineer: $83,000-$195,000
• Data analyst: $50,000-$128,000
• Data scientist: $76,000-$195,000