My robot was working before, but it suddenly stopped working. What should I do?

There are two types of robots on Browse AI: Prebuilt Robots and Custom Robots

Prebuilt Robots are maintained by automation specialists and if you come across an issue with them, you should contact our support team immediately to look into them.

Custom Robots, on the other hand, can be created by you or your team members on any of the billions of websites out there.

One general challenge with web automation and scraping is that websites change. The most popular websites change all the time, sometimes several times a day, and Browse AI is designed to automatically adapt to changes. This approach works most of the time: our data shows an average custom robot can adapt to website changes on its own for 6+ months.

In the rare cases when a robot is working successfully and then it stops working (meaning its tasks fail or extract the wrong data), all you need to do is to re-train the robot. You can do so by going to the robot's Settings tab and clicking on the "Re-train robot" button. Training a robot only takes a couple of minutes.

If the robot logs in on your behalf, you should also try pressing "Update Session Cookies" on the robot Settings page. The issue may just be that the cookies have expired and the robot is unable to log in on your behalf.

If you keep re-training the robot quite frequently and it works successfully only for a little while, you can contact our support and see what else you can do to make the robot more resilient. It may be something you can do differently when training the robot.