Data Quality & Fitness
Accuracy, completeness, timeliness, consistency, and fitness for the intended use.
Home ARDDThe AI-Ready Data
ARDD
The AI-Ready Data
Diagnostic (ARDD).
A maturity assessment model for the data foundation of enterprise AI, developed by Sami Tayara through academic research. This overview introduces ARDD's nine foundation dimensions and its two contextual dimensions.
ARDD / Foundation tier
Nine dimensions of readiness.
Lineage, Provenance & Auditability
Traceability of data, audit logs, and evidence of where an output came from.
Meaning & Semantics
Owned and connected metadata, business definitions, and semantic models.
Sufficiency & Representativeness
Enough data for the intended use, representative of the population the system acts on.
Ownership, Stewardship & Decision Rights
Accountable owners, funded stewardship, and clear decision rights.
Rights & Permissible Use
The purposes for which data may lawfully and contractually be used.
Access, Security & Privacy
Least-privilege access for people and agents, and protection of sensitive data.
Architecture & Accessibility
Discoverable data products, integration, pipelines, and access for applications and agents.
Sovereignty & Residency
Data residency, cross-border flows, and reliance on vendors and compute.
ARDD / Contextual tier
Two contextual dimensions.
ARDD reports these alongside the foundation assessment. They are excluded from the foundation index and floor.
Activation Readiness
The path from governed data to prioritized AI use cases and business value.
Delegated Authority
The extent of authority delegated to AI systems and the governance of that authority.
The maturity scale
From absent capability to continuous improvement.
0No Capability
1Initial
2Repeatable
3Defined
4Managed
5Optimized
Levels are cumulative. A level counts only where its criteria and every lower level's criteria are met.
ARDD uses the weakest foundation dimension to set each use case's floor, and the lowest use-case floor to set the organization's band. The averaged foundation index provides a secondary view of direction. The published paper sets out the full measurement rules.
Two complementary frameworks
Readiness beneath the trust wing.
ARDD is a separate authored work that assesses the data foundation. Its findings can inform AI-Ready Data, the foundation of TWAF's trust wing. TWAF assesses the wider initiative across value and trust. Data and AI work progress in parallel, with use-case requirements informing data priorities.
Framework ↗