Check a dataset before you trust it.

Missing values, duplicate rows and keys, malformed e-mails, impossible dates, out-of-range numbers, outliers and orphaned foreign keys — scored and explained. Everything runs in this browser tab; your file is never uploaded.

1 Choose a file

Drop a file here or click to browse CSV, TSV, Excel, JSON, JSONL or Parquet

Starting the Python engine…

2 Set the rules

Import rules

Starter rules were suggested from your column names and values — review them. The exported file works unchanged with the command-line tool (data-quality data.csv -c rules.json).

Column	Type	Required	Unique	Date	Format	Min	Max	Must exist in

Flag columns missing more than % Date format Allow future dates Outliers

Add a reference table Optional: add e.g. a customers file to check that every order's customer exists.

Edit rules as JSON

3 Results

Findings

Column profile

Column	Type	Missing	Distinct	Range	Most common