Launching today
DataCleanist
80% of ML is data cleaning. We do that part.
3 followers
80% of ML is data cleaning. We do that part.
3 followers
DataCleanist checks your dataset with real code before any AI touches it. Upload a CSV or Excel file and get an interactive, column-by-column report: missing values, outliers, duplicates, correlations, distributions, VIF, feature importance, statistical tests, categorical analysis, MCAR/MAR and target leakage. Then ask questions in plain English and let it clean the file. The AI answers from that report, so it cannot invent a number. Your data is never used to train models.
Hey Product Hunt, Anas here.
I built DataCleanist because I got tired of doing the same data-cleaning setup every time a new dataset landed: checking types, hunting for missing values, spotting outliers, removing duplicates, and figuring out whether a column was quietly leaking the target.
DataCleanist does that first pass automatically for CSV and Excel files, then shows you exactly what it found and why each issue matters. You can also chat with the report, generate the Python cleaning code, and keep every cleaned version without touching the original.
I’d genuinely love to know what you’d want it to catch next.
If there’s a data-quality problem you regularly discover only after training a model or starting an analysis, tell me about it.