Beyond Basic CSV Validation
Most CSV tools only check whether a file parses - this one looks at whether the data makes sense. For each column, it detects the dominant type (mostly numbers, or mostly text) by checking what at least 60% of the non-blank values look like, then flags two kinds of problems: blank cells, and values that don't match their column's type - like a stray "N/A" or "unknown" sitting in an otherwise all-numeric Salary column.
For every blank cell, it suggests a fill based on the rest of that column: the column average for numeric columns, or the most frequently repeated value for text columns - visible directly in the highlighted preview, with a one-click copy of the filled-in CSV.
How Suggestions Are Chosen
Frequently Asked Questions
No - parsing, analysis, and gap-filling suggestions are all computed entirely in your browser using plain JavaScript. Your data is never uploaded or sent to a server.
It checks every non-blank value in the column and treats it as a numeric column if at least 60% of them parse as valid numbers - a small number of text exceptions (like "unknown" or "N/A") in an otherwise numeric column will be flagged as mismatches rather than changing the column's overall type.
No - a column average or most-common-value is a reasonable statistical guess, not a fact about the specific missing record. Always review suggested fills before using them, especially for anything that matters (like financial data).
No - mismatched values (like text in a numeric column) are only highlighted for your attention, never automatically changed, since there's no safe way to guess what the correct value should have been.