KB article
Tooling Interfaces: SQL, DAX, and the Translation Layer
Different tools expose different query layers; AI must align with them.
arf-kbai-readiness-interoperabilitysemantic-compatibilitysemantic-contractdeterministic-query
TL;DR
- AI answers depend on the query layer.
- Translation layers must preserve meaning.
The problem
- SQL and DAX can yield different results if semantics differ.
- AI might query the wrong layer.
Why it matters
- Cross‑tool consistency is required for trust.
- Interoperability reduces maintenance cost.
Symptoms
- Results differ between SQL queries and Power BI visuals.
- AI answers are inconsistent across tools.
Root causes
- Semantic logic only exists in DAX.
- Translation between layers is unclear.
What good looks like
- Semantic logic shared or mirrored across layers.
- Clear documentation of which layer is authoritative.
How to fix
- Identify canonical layer for each metric.
- Document translation rules.
- Validate consistency across tools.
Pitfalls
- Assuming SQL and DAX are interchangeable.
- Ignoring semantic differences in calculations.
Checklist
- Authoritative layer defined.
- Translation rules documented.
- Cross‑tool tests pass.
Framework placement
Primary ARF layer: AI Readiness & Interoperability. Diagnostic bridge: data-movement-reliability, semantic-reliability, execution-reliability, change-reliability.