Quickstart

Go to the dedupe tool. You'll see a drop zone with two buttons: Choose CSV and Try a sample. If this is your first time, click Try a sample to run a built-in messy customer list and see the whole flow before you upload your own data.

Drag a CSV onto the drop zone, or click Choose CSV and pick a file. Any list with duplicate rows works — contacts, customers, members, event attendees, accounts. The free tier cleans up to 1,000 rows per run; larger files are a Pro feature.

The matching engine works out the rules from your columns — no setup, no threshold tuning — and groups rows that refer to the same real-world entity. This usually takes a few seconds.

Each duplicate group gets a verdict from the review agent:

• Confirmed — the rows really are the same entity.
• Uncertain — the evidence is thin or mixed; a human should look.
• Rejected — the agent thinks the engine over-merged two different things.

Rejected and Uncertain groups are floated to the top under Needs review, each with a one-line explanation. Click any group to expand its rows and see which one is kept and which are dropped.

Click Download clean CSV. One survivor row is kept per duplicate group, plus every row that wasn't a duplicate. The file is assembled in your browser and saved to your machine as yourfile-cleaned.csv.