See how an Octavus Agent controls its computer with the built-in browser and by clicking specific areas on the desktop.
Two ways Agents drive the computer
Agents have a few different tools that make it effective for them to work on their computer. Two primary methods are:
- Browser - the Agent controls the built-in browser to navigate, type into fields, and search.
- Desktop - the Agent maps areas on the computer screen and clicks specific locations.
A task that needs both, such as downloading an image and then searching with that image, shows both methods.
Drive the built-in browser
- Give the Agent a task that needs the browser, such as searching for an image.
- Watch the thread while the Agent opens the built-in browser.
- Follow it as it navigates, types into fields, and searches.
Anything that happens in the browser can be done this way.
Click specific areas on the desktop
Some work needs the desktop, not just the browser. For example, a CAPTCHA may appear while the Agent is using the browser.
- The Agent can pull up a worker to help.
- It maps locations on the computer screen.
- It clicks those specific areas to complete the challenge.
CAPTCHAs are designed to stop old-school bots. Octavus Agents are knowledge workers, so they can complete this kind of challenge and keep going.
Verify
The thread shows browser steps such as opening a tab, typing, and searching. When the Agent needs the desktop, you see it map screen locations and click them, then continue the original task.