rasti.id.ir

Portfolio12

Upwork · E-commerce · 2024

Scrape → Sheets → Woo

Supplier catalogs change faster than merchants can type. Blind auto-import destroys trust; full manual can’t keep up. I built the middle path: scrape, AI-format, humans approve in Sheets, PHP pushes to Woo, Telegram watches every hop.

My role Pipeline designer — scrape, AI normalize, Sheets control plane, Woo import plugins, Telegram logging.

DataAutomationWooCommerce

Challenge

Ecommerce clients on Upwork needed catalog sync without publishing garbage. Reviewers lived in Sheets. Failures had to surface in Telegram — not a log file nobody opens.

What I did

Browserless for capture, AI for normalization, Google Sheets as the approve/reject control plane, WooCommerce import plugins for the push, Telegram for ops visibility. Same philosophy as the Bondi/TCD migration tooling — structure first, publish second.

Outcome

Catalog velocity with a human brake — approved rows only, failures loud enough to fix.

Sheets as the control plane

Non-developers already trust spreadsheets. Making Sheets the approval gate meant the pipeline fit the team instead of forcing a new admin UI.

Telegram is the pager

If a scrape dies or an import fails, it has to land where people already watch. Silent pipelines are how bad data ships.

What stands out

Scrape + AI format

Browserless capture and AI cleanup before a human opens the sheet.

Approve then push

Only reviewed rows hit WooCommerce.

Telegram trail

Success and failure land in the ops channel.

Stack

Browserlessn8nPythonGoogle SheetsPHPWooCommerceTelegram

Roles

DataAutoAIWP

Tags

Browserlessn8nPythonSheetsWoo

More case studies

More of my builds across WordPress, automation, data, and product.

All projects →