Automated AI Web Scraper

This workflow uses the Zyte API to automatically detect and extract structured data from E-commerce sites, Articles, Job Boards, and Search Engine Results (SERP) - no custom CSS selectors required.

It features a robust "Two-Phase Architecture" (Crawler + Scraper) that handles pagination loops, error retries, and data aggregation automatically, ensuring you get a clean CSV export even for large sites with thousands of pages.

If you prefer to use your own parsing logic and just need raw data, it provides a "Manual Mode" for that capability as well.

Supported Modes

• E-commerce / Product: Extract prices, images, SKUs, and availability.
• Articles / News / Forums: Extract headlines, body text, authors, and dates.
• Job Boards / Postings: Extract salaries, locations, and descriptions.
• SERP (Search Engine Results): Extract search rankings, organic results, and snippets.
• General Scraping: Get raw BrowserHtml, HTTP Response codes, Network API traffic, or Screenshots to parse yourself.

How it works

• Input: You enter a URL and choose a goal (e.g., "Scrape all pages") via a user-friendly form.
• Smart Routing: A logic engine automatically configures the correct extraction model for the target website.
• Two-Phase Extraction: (Active only for "Scrape all pages") Phase 1 maps out all available URLs (Crawling), and Phase 2 extracts the rich data (Scraping), filtering out errors before saving to CSV.

Set up steps

1. Get your API Key: You need a free Zyte API key to run the AI extraction. Get it here.
2. Run: Open the Form view, paste your key, select your target website, and hit Submit.
3. Export: The workflow will process the data and output a downloadable CSV file.

Resources

• Zyte API Documentation
• Get Help (with API errors & extraction logic)