Overview
Open data is only useful if it's clean. This runbook profiles each dataset, checks it against expectations and flags issues before publication.
Steps
- Profile every column (types, ranges, nulls, uniqueness).
- AI proposes validation rules from the data + the dataset's description.
- Flag rows/columns that violate expectations with a clear reason.
- Produce a quality report the data owner signs off before publishing.
Outcome
Higher-trust open data and fewer post-publication corrections.