Data profiling
Measure what each source really contains (completeness, granularity, duplicates, value ranges and quality problems) before the estimate and the design rely on it.
- Use when
- Data warehouse, BI or migration from sources nobody has measured yet.
- Avoid when
- Clean, well-known sources that were already profiled.
- Source
- DAMA-DMBOK: Data Quality Management
- Level
- B · established practice
For agents · MCP
explain_node({ "id": "t.data_profiling" })