Data quality, automated
No LLM needed. Rules come from the data's own statistics. Free to run, same result every time.
Everything runs locally in your browser with Pyodide. Your file is never uploaded.
Run the rules on fresh data. The sample batch has 10 planted problems: a duplicate ID, a missing and a malformed email, an impossible date, a stray space, a US zip in a Canadian postal code column, a new plan value, negative values and an outlier.