From Publisher Catalog to Live Product Pages, Automatically

The problem

Listing new books was a marathon of manual steps: scrape the publisher’s catalog, rewrite each description, track which titles were already listed in a spreadsheet, then enter every product into two separate storefronts — one of them behind two-factor authentication. Hours of repetitive work for every publisher batch.

The build

We built a five-module pipeline that takes a single publisher URL and runs the whole way to live product pages.

It scrapes the publisher’s catalog — every page, every product URL, full details and images — using stealth browser sessions paired with TLS-fingerprinted requests so the source site doesn’t block it. Each book’s description is then formatted by AI into clean HTML, meta text, and tags, and recorded in Google Sheets, which acts as the source of truth so already-listed titles are skipped.

From there it publishes automatically into both storefronts, generating valid authenticator codes on the fly to clear the two-factor login on the second one. A dashboard sits on top — run status, logs, history, and scheduled triggers — so the client operates the whole system without touching code.

The result

A publisher URL goes in; fully formatted, deduplicated, published products come out across both storefronts. This was the client’s third automation project with us — after a daily AI content pipeline and a POS import bot.