AI Data Cleaning Workflow

Help me profile this dataset. Columns: [list columns with types]. Generate a data quality checklist: what to check for each column type, common issues to look for, and questions about the data I should answer.

I have missing values in these columns: [list columns with % missing]. For each, recommend: whether to drop rows, impute (and with what), or keep as-is. Consider the data type and what the column represents.

Help me identify and handle duplicates. My data: [describe]. What defines a duplicate? [your criteria]. Write code to: find exact duplicates, find fuzzy duplicates, and decide which to keep based on [criteria].

Standardize these columns: [list columns with issues like "dates in multiple formats" or "inconsistent category names"]. Write code to clean each one and document the transformations.

Create data validation rules for this dataset. For each column: valid ranges, expected patterns, logical relationships between columns, and how to flag violations. Generate validation code.

Document the data cleaning I performed: [list transformations]. Create: a summary of changes made, rationale for each decision, before/after statistics, and code for reproducing the cleaning.