Data quality is often described as a technical problem, but most persistent quality issues are operating problems. They come from unclear ownership, weak validation, mismatched definitions, or workflows that never create the right record in the first place.
Engineering can detect and reduce defects, but the business process has to explain what good data means and who is responsible for it.
For AI-assisted operations, this matters more. Agents need reliable inputs and clear fallback paths when the underlying records are incomplete or contradictory.