Web scraping kind of sucks, even the no-code tools. No one wants to define schemas or click on elements on a page over and over again.
What if there was a different way? A better way? A lazier way. . .
My latest late-night weekend side project. Here's how it works:
- Define your fields
- Submit a link, any link
- The program extracts all HTML content
- The program absorbs your fields and uses GPT-3.5 to extract accurate data
- Rinse & repeat on different sites without defining HTML elements!
Why Morphix? "Morph" refers to the process of transformation or change, while "fix" implies stability or resolution. Together, Morphix transforms your data into a stable, readable state.
- Give a field as much detail as possible, like "Job Title" instead of just "Name" or "h1"
- Add all of the links you want to scrape, then click collect, and export together.
- Avoid deleting fields after you've generated data, or just be ready to click "collect" again after!