Loading Datio workspace…
Preparing the browser-side engine
Loading, please wait…Preparing the browser-side engine
Loading, please wait…Deep quality analysis of your dataset.
Load a CSV file to start profiling.
The data profiler is a data profiling tool and data quality checker for any CSV-style file. It reports row and column counts, per-column types (numeric, text, date, boolean, mixed, empty), unique and missing counts with samples, plus duplicate-row and missing-cell totals. Six computed scores — completeness, consistency, validity, uniqueness, accuracy, and timeliness — roll up into an overall quality score with plain-language findings.
Useful before any cleanup or presentation: analysts auditing an extract, operators checking a list for blanks and repeats, and students learning what data quality actually means.
Column "email" is missing 12 values (6.0%).
Duplicate rows detected: 3.
→ clean blanks and repeats, then re-profile to confirm zero.Profile first, then act: clean duplicates and blanks, narrow rows, and re-profile to confirm the fix before visualizing. CSV Cleaner, CSV Viewer, Visualizer.
Data profiling means auditing a dataset's shape and quality before using it: column types, unique and missing counts, duplicates, and consistency. Datio's Data Profiler computes all of this locally and rolls six scores — completeness, consistency, validity, uniqueness, accuracy, timeliness — into one overall quality score with plain-language findings.
Load the file in Datio's Data Profiler and read the overall score first. Then check the duplicate-row and missing-cell totals, scan the Findings list for empty columns or mixed types, and open Column Analysis for per-column unique counts, null percentages, and sample values. Export the profiler report as CSV to keep the audit.
A column is typed when roughly 80% of its non-empty values agree (numeric, text, date with - / or :, or boolean). When a column mixes substantial numbers and text, it is flagged as mixed — usually a sign of stray text in a numeric column that cleaning or filtering should fix.
The Viewer is for reading: sorting, searching, and eyeballing rows. The Profiler is for auditing: type detection, missing-value analysis, duplicate counts, and quality scores. View to understand, profile to judge — then clean, filter, or chart based on what you find. CSV Viewer →
The profiler audits quality — it does not fix data itself, and there is no automatic imputation. Timeliness is derived from date columns when present; files without dates fall back to consistency-based scoring.