Catalog Cleaner

Sample catalog · Amazon flat file audit

Removed 0
Flagged 0
Clean 0

Rules

0 rows read

Removal rules run first. The first rule that matches sets the decision.

Remove

  • R2Remove0

    Image count below images

  • R4Remove0

    Price is 0, or above USD

  • R6Remove0

    The same GTIN occurs again: remove the later row

  • R7Remove0

    Category is not in the allowed list

Flag for review

  • R1Flag0

    Brand is missing

  • R3Flag0

    Title longer than characters, or more than % capitals

  • R5Flag0

    Title has a banned word: best, free shipping, #1, hot sale

  • R8Flag0

    AI check: title has no brand word, or the price is far from the category median sample

Allowed categories
DressesTopsBottomsOuterwearBagsBeautyShoesJewelryWatches

Sample AI output. In production a model call replaces this file.

Scale test

250,000 rows, made and checked in the browser

Idle

Rows
—
Removed
—
Flagged
—
Clean
—
Run time
—
Rows per second
—

The scale test runs the seven deterministic rules (R1 to R7). R8 needs a model call for each item, so this demo does not run it at scale. In production, clean rows go to the model in batches.

Review queue

Loading sample catalog…

SKU Title Category Price Images Decision Reason

SKU

Item