Skip to main content
All CollectionsBest practices
How to perform a custom search using Bulk Run
How to perform a custom search using Bulk Run

Extract data from multiple sources by combining Bulk Run with custom input parameters to automate searches across many URLs efficiently.

Nick Simard avatar
Written by Nick Simard
Updated this week

What you'll learn

This guide explains how to combine two powerful Browse AI features: Bulk Run and custom input parameters. Together, they allow you to run a single robot across multiple URLs and search terms, automating data extraction at scale and saving hours of manual work.

Understanding the process

When you train a custom robot, any text inputs you interact with (like search boxes) can be set as input parameters. Later, you can adjust these parameters for different tasks without creating new robots. Using Bulk Run, you can upload a CSV file containing multiple search terms or URLs to extract data from many sources in one operation.

Step 1. Build your robot with input parameters

Create a custom robot for your chosen Origin URL and manually demonstrate the search you want it to perform, and then proceed to capture your desired data:

  1. Go to your dashboard and click on Build New Robot

  2. Enter the URL of the webpage you want to extract data from

  3. Click on Start Training Robot

  4. Train your robot by performing necessary actions on the webpage (searching, clicking, typing)

  5. Click on the capture button and choose between capturing a list or capturing text

  6. After capturing all necessary data, click Finish recording

  7. Name your robot and click Save

This will provide a template for the robot to follow. Browse AI robots are built to interact with websites just like you would. They can perform searches, clicks, and navigate between pages.

After the robot has been built, you should have another parameter apart from your Origin URL and data limit.

Step 2. Prepare your CSV file for Bulk Run

After building your robot with input parameters:

  1. Click on the Run Task tab and select the Bulk run tasks button in the lower-right corner

  2. Download the Sample CSV file provided with your newly trained robot

  3. Modify the CSV file by adding or editing columns for your search terms or parameters

    1. Each row should represent a different search you want to perform

    2. Ensure your CSV has a header row with column names that match your input parameters

  4. Save your CSV file with the updated information​

Step 3. Execute the Bulk Run

  1. Return to the Run Task tab of your robot's dashboard

  2. Click the Bulk Run Tasks button

  3. Upload your modified CSV file

  4. Select the column headers that correspond to your input parameters

  5. Optionally, set up an integration to send the extracted data to another tool

  6. Click "Run [x] Tasks" to begin the process (X amount is the number of rows of data you are about to extract)

The search keyword data point can vary in type - they can be names, numbers, URLs, or other text values.

Important considerations

  • Ensure you have enough credits for your bulk run. To calculate requirements, check How are credits calculated?

  • If you don't have enough credits, you can run the task with a lower data limit

  • Browse AI will automatically run your robot on each search term in your CSV file, extracting the data into a compiled output

  • You can view and analyze your results in the "Tables" section (See: How to use data tables)

Troubleshooting tips

  • If your bulk run fails, check that your CSV formatting is correct with no special characters in the header row

  • Verify that your robot is working correctly by running a single test before attempting a bulk run

  • For search parameters that include special characters, ensure they're properly formatted in your CSV file

By combining Bulk Run with custom input parameters, you can transform a single robot into a powerful data extraction tool capable of handling thousands of different searches automatically.

Did this answer your question?