How to build a robot with Robot Studio?

Table of Contents:

How to access Future Labs

Using the Robot Studio



Browse AI has an alternative solution to Chrome extension called Robot Studio. The key difference? No need for a browser extension, therefore you can use it across various browsers other than Chrome!✨ Just a heads up, it's in beta, so we're fine-tuning things to ensure maximum stability. Our beta features can be found in the Future Labs. Let's take a look at how to access this feature and how to utilise the Robot Studio:


How to access Future Labs:

  1. On your Browse AI Dashboard, click on your profile picture in the top-right corner, then choose Future Labs like so:

    4f509372e568359fb1994b1feaccb944.png

  2. Next, toggle the switch, “Train custom robots in the New Robot Studio”:


  3. Congrats! It’s officially enabled now.

Using the Robot Studio:

  1. Once you’ve enabled the Robot Studio on the Future Labs, go ahead and build a new robot by going to your Dashboard > Build a New Robot:


  2. Or if you’ve yet to build a robot, feel free to pick between “Monitor Site Changes” or “Extract Structured Data” based on your preferences:



  3. Next, input your desired link on the Origin URL text field, then select “Try New Robot Studio (Beta)”, like so:

    Or you might see something like this instead:


  4. A new window will pop, something like this:

  5. Once the webpage has loaded, you can extract the data you're after, either From a list or by Just Text


Here’s a sample screenshot capturing a data set From a list

    1. When capturing a list, hover over the container where your data is, and the robot will smartly detect it. Then, choose the specific data points you want to extract, like title or name, description, and even the Image logo URL. You can press the Escape key to cancel the capture and correct it. 

      Once you’re finished, press the Enter key to label your data points. You can name them anything you like: 

      After adding those labels to your data items, a window opens up, giving you a chance to review how your data is shaping up. Plus, we've got three sections you can tweak: 

      1. First, from top to bottom, the name of your list. You can name it anything you like. 
      2. Next up, at the bottom-left, you can decide on the amount of rows you want to extract. 
      3. Third, at the bottom-right, pick the pagination type that matches the website you’re on. In this example, the website loads more content as you scroll down.
      4. Next, go ahead and choose the "Save Captured List" button.

        Once everything looks good, hit the "Finish Recording" button boxed in yellow. But if you want to capture another set of data, you can go ahead and select Capture Text, boxed in turquoise:

      5. If you're up to extract another dataset, you can either capture another list, or capture a text. 🙌

      6. In the next example, let's capture a dataset by Just Text. You can select more specific data points here, such as header, body, and so on:

      7. Capture those texts you need, and when you’re ready to label them, just hit the Enter key:


      8. Then, a new tab will appear, allowing you to switch between both your new data set and the first data set. Once you’re okay with the data, just hit “Save Captured Texts”, boxed in red:

      9. After that, you can either extract another dataset by clicking on "Capture Texts", or you can click on Finish in the bottom-right corner to start finalizing your data capture: 

        Once you click on the "Finish" button, you’ll be prompted to name your robot. Feel free to name it anything memorable, then click on Save:

        Afterward, you'll be able to look at your extracted data with the final screen and the outlined steps beside it. Here’s a sample of the last data:

      • Just a quick heads up! You can download a copy of your extracted data in CSV or JSON format.
    2. If everything checks out, scroll down to click on “Yes, looks good,” like so: 

      And there you have it! Fantastic job navigating the Robot Studio. 🎉

Did this answer your question? Thanks for the feedback There was a problem submitting your feedback. Please try again later.

Still need help? Contact Us Contact Us