The Origin URL serves two important purposes: it's the starting point when you train a new robot, and it becomes a flexible parameter for scaling that robot across similar pages.
Origin URL during robot training
When you click Build new robot and pick Extract structured data, the Origin URL is on the next screen, before you pick Table Studio or Robot Studio. This is where you tell Browse AI which webpage to use for training.
What to enter: Enter the exact webpage URL where your target data is currently displayed. Choose a page that gets you as close as possible to the data you want to extract.
Good Origin URL choices for training:
A specific product page:
https://store.com/products/laptop-model-xAn individual company profile:
https://directory.com/companies/airbnbA particular search results page:
https://jobs.com/search?role=engineer&location=nycA detailed listing page:
https://marketplace.com/listings/vintage-guitar-123
Starting close to your target data makes training faster and creates more reliable robots.
Origin URL as an input parameter
After your robot is saved, the Origin URL becomes an input parameter that you can adjust to reuse the same robot across multiple similar pages.
How to scale your robot
Open your robot and go to the Run Task tab.
Under Run a task now, replace the Origin URL with a different page that has the same layout. To keep it for future runs, click Save as default.
Click Run a Task.
Scaling examples: If you trained a robot on one company profile, you can extract data from other company profiles by changing the Origin URL:
Original training URL:
https://directory.com/companies/airbnbScale to:
https://directory.com/companies/uberScale to:
https://directory.com/companies/stripe
In a Table Studio robot's Tables tab, you can also click New Row and enter another Origin URL, or Import CSV to add a list of them. Each row runs the same robot on its own page.
Ways to scale using Origin URL: