I built a small data-cleaning workflow around one simple rule: bad rows should never silently dis...

I built a small data-cleaning workflow around one simple rule: bad rows should never silently dis...I built a small data-cleaning workflow around one simple rule: bad rows should never silently dis...

The network for creativity

Join 1.25M professional creatives like you

Connect with clients, get discovered, and run your business 100% commission-free

Creatives on Contra have earned over $150M and we are just getting started

Back to feedPost

Julian Vance

• 5h

I built a small data-cleaning workflow around one simple rule: bad rows should never silently disappear.

Synthetic demo results:

• 20 raw order records • 16 clean rows • 2 duplicates removed • 2 invalid rows quarantined • $3,690 paid revenue reconciled

The output includes a clean CSV, a rejected-row file, an audit log, and an Excel dashboard. Dates, names, emails, regions, products, and payment statuses are standardized.

The key design choice is quarantine + reasons, not just dropping bad rows and exporting. That lets a reviewer trace exactly what changed and decide whether rejected records should be repaired.

Portfolio: https://contra.com/p/WPLXml6R-demo-automated-sales-data-cleaning-and-excel-reporting

For recurring exports, what needs the most manual review in your workflow: duplicates, inconsistent labels, or invalid amounts and dates?

Data Analysis Microsoft Excel Python

Advik Rathee

• Jul 14

Built this one for an e-commerce customer behavior project - 18,000+ customers, segmented using RFM (Recency, Frequency, Monetary) analysis.

The interesting part wasn't the segmentation itself, it was what it revealed: nearly 22% of customers were "at risk" of churning, and most businesses wouldn't know that until those customers had already stopped buying.

This is the kind of insight that's usually sitting in your order data already, it just needs the right lens. Happy to take a look at your numbers if you're curious what's hiding in yours.

Tools used: Power BI, SQL, Python (Pandas, Scikit-learn for clustering)