Rewired, 2nd Edition

CHAPTER 25No data architecture, no AI advantage

Data architecture is the set of connected capabilities that determine whether the enterprise can find, trust, and serve data to analytical and AI systems at scale. A modern data and AI architecture has shared capabilities that work together as a system.

In this chapter, we refer to five core data and AI capabilities that show up, in different forms, in every successful architecture.

A modern data architecture

A modern data architecture spans eight tightly connected capability areas, each playing a distinct role in turning raw data into trusted, usable assets for the business. Starting from the foundation:

1. Data ingestion. This capability brings data into the enterprise. It covers the services that collect data from transactional systems, events, documents, and external sources, while handling sensitive information appropriately. Its role is to ensure that data arrives reliably, securely, and in a form the organization can depend on.
2. Data platform. The data platform provides the durable foundation for storing, processing, and transforming data at scale. It includes processing services that clean, enrich, and prepare raw data—structured and unstructured—for analytics, AI, and retrieval, as well as the storage services that manage raw and refined data reliably.
3. Semantic layer. The semantic layer makes enterprise data understandable to both humans and machines. It includes shared business definitions, metrics, and models that establish ...