COMPEL Glossary / data-architecture
Data Architecture
Data architecture is the design of how data is collected, ingested, stored, organized, integrated, transformed, governed, and made available across an enterprise to support AI capabilities, analytics, and business operations.
What this means in practice
It encompasses technology choices (data lakes, warehouses, lakehouses, data meshes), organizational decisions (centralized versus federated data ownership), and governance mechanisms (quality standards, access controls, lineage tracking). For organizations pursuing AI transformation, data architecture is often the greatest enabler or the greatest bottleneck because AI models are fundamentally dependent on the quality, availability, and accessibility of data. In COMPEL, data architecture is a core Technology pillar domain assessed during Calibrate and designed during Model, with enterprise-scale architecture patterns covered in Module 3.3, Article 3.
Why it matters
Data architecture determines whether AI initiatives have the foundation they need to succeed or are doomed to struggle with fragmented, inaccessible information. Organizations with poor data architecture spend up to 80% of AI project time on data wrangling instead of model development. A well-designed data architecture accelerates time-to-value for every AI use case and creates compounding returns as new initiatives build on shared infrastructure.
How COMPEL uses it
During the Calibrate stage, COMPEL assesses data architecture maturity under the Technology pillar (Domain 10), identifying whether current infrastructure supports or bottlenecks AI ambitions. The Model stage then designs the target data architecture as part of the enterprise AI platform blueprint. Module 3.3, Article 3 provides advanced architecture patterns including data mesh and lakehouse designs for AITGP-level practitioners.
Related articles in the Body of Knowledge
Related Terms
Other glossary terms mentioned in this entry's definition and context.