Listing Quality Auditor
A product-data quality auditor that reports what’s wrong and proves it, without touching your catalogue. Rules live as data, so changing one takes an edit rather than a release. Zero dependencies, 13 tests.
The problem
Large e-commerce catalogues drift. Titles pick up typos, variants start contradicting the listing they belong to, specifications stop matching the item, prices go stale against the price list. At tens of thousands of listings nobody can find these by hand — and a tool that silently corrects them is worse than the drift, because now the errors are invisible.
What we built
An auditor that reads a catalogue export, checks every row, and produces a prioritised report. It never edits the source and never fixes anything: a person decides what changes. Every finding comes with the evidence behind it, so nothing has to be taken on trust. When a check cannot be verified — a missing field, an incomplete row — it says so and names what it needed, rather than passing quietly. Auditors that skip what they cannot judge produce clean reports and false confidence. The rules live as versioned data rather than in code, so changing one is an edit and a version bump, not a development cycle. Built against catalogues in the tens of thousands of listings, it runs on zero dependencies and is covered by 13 tests.
Start a project
Open a ticket
Tell us what you’re building. We reply within 24 hours with honest thoughts on scope, timeline, and price.