Skip to main content

Data Architecture for Software Architects

Data architecture is choosing where and how data is stored and served. It spans relational, document, graph, columnar and time-series stores, caching layers and read models, analytics pipelines, and search and vector stores, each matched to a query pattern rather than used for everything.

Stage: Data and Storage · Part of the Software Architect roadmap · Last reviewed 19 November 2026

What to learn in Data Architecture

Data Architecture interview questions

How do you choose between SQL and NoSQL?

Start from access patterns, consistency needs and data shape. Relational stores suit structured data and transactions with joins, while document, key-value or wide-column stores suit flexible schemas and horizontal scale. Many systems use several.

What are the main cache invalidation strategies?

Common approaches are time-to-live expiry, write-through updates, explicit invalidation on change, and event-driven invalidation. Each trades freshness against complexity and load. Stale data tolerance for the use case should decide the choice.

Where Data Architecture fits in the Software Architect roadmap

Data Architecture belongs to the Data and Storage stage (advanced level), which takes about 5 hours. Pick the right store.

Back to the full Software Architect roadmap