How UiPath Enables End-to-End Automation on macOS (Full Transcript)

[00:00:00] Speaker 1: When it comes to enterprise-level automations like creating agentic processes or building RPA workflows, Mac users historically haven't gotten a lot of love. Many automation tools and applications are Windows-only or require workarounds on Mac, like using virtual machines or lite versions of tools. But now, UiPath has changed things by enabling full end-to-end automation directly on macOS using Studio Web in your browser and their UiPath Assistant for Mac. So you can now fully and natively automate everything from personal productivity tasks to enterprise-grade processes that involve multiple apps, systems, and steps. You can do this entirely on your Mac, and you have native UI automation for any type of Mac application. So you can automate tasks in apps like Pages, Numbers, Notes, Mail, Safari, and even cross-platform apps like Excel or Word. Hi, I'm Garrick, and in this video, we're going to take a look at some example automations that have been set up and run entirely through UiPath Studio Web on a Mac. To get started, point your browser at cloud.uipath.com. The first time you do this, you'll be prompted to sign in to your UiPath account, so if you haven't created one yet, you can set one up for free from here. Once you've signed in, you'll see your dashboard. Over here on the right, click Download Center, and on the Downloads page, you'll find Assistant for macOS. Now, you'll notice that UiPath offers both Intel and Apple Silicon versions of the UiPath Assistant for Mac. But it's important to know that only the Apple Silicon version supports UI automation for native Mac desktop apps, which is what we'll be looking at in this video. If you're using an Intel-based Mac, Studio Web can still run cross-platform automations and web UI automations, but it won't support desktop app automations like the ones I'll be demonstrating, so just bear that in mind. I am on an Apple Silicon Mac, and I've already downloaded and installed the Assistant on my Mac, which you can see here. Just make sure you sign into it with the same account you signed into UiPath with. And there's actually a fair amount you can do in the app, like try out some pre-built automations or browse the Marketplace for other automations. But in this case, we just need the UiPath Assistant running so it can help with the automations we build and run from our web browser. Now the first time you open the UiPath Assistant, you'll be prompted to allow it certain permissions in system settings, like accessibility and screen recording. So make sure you allow it access to those settings, since that's what allows UiPath to securely interact with your screen and local Mac applications. But now let's go back to the web browser where we're signed into my UiPath account. I'll click Studio to go into Studio Web. Studio Web is UiPath's browser-based, AI-native build environment, which means there's nothing else to install on your Mac to start building applications. And here you'll see any automations you've already created. I have a few here that I've been working on, so let's take a look at some examples. Here's one called Full Name CSV Workflow. So if you've ever worked with UiPath on Windows, this should look pretty familiar to you, only this is entirely in my Mac's browser. And this is a simple workflow that reads a CSV file on my desktop, looks at the content of each row, and generates a new file with additional data. So here's what the file looks like in Excel, and you can see it has just two columns labeled First Name and Last Name, with some names below. So this automation is set up to look at that file, read the contents, and generate a new CSV called Names Underscore Full, which will include a third column called Full Name, containing the combined first and last names in each row. So again, this workflow will begin troubleshooting directly in the web browser, and then continue running locally with the help of UiPath Assistant for Mac. So again, it's important that my assistant is currently running, allowing Studio Web automations to interact with the local files and apps on macOS. So let's run it here and see if it works. I'll click Debug on local machine. And as it's running, we can see a log of exactly what it's doing, and we're seeing the successful label next to each step of the workflow as it runs through it. And you can see here that it tells us that the execution has ended. And now if we look on the desktop, here's the new NamesFull.csv file it generated. And I'll open that up in Excel. And we can see that it has successfully created the new Full Name column with the combined names from each row. So that's one example of an automation you can run in Studio Web on your Mac. Now a workflow like this one does assume that you have some familiarity with how UiPath automations are built. And if you want a deeper walkthrough of that process, Kevin has a great video on this channel that I'll link to in the description. But the good news is that UiPath also includes tools that make UI automation much more approachable, where you can simply indicate what your robot clicks and where without having to write any code. So for example, I have this document open in the Pages app. Let's say I have to frequently go through a process where I need to add a legal disclaimer page to company documents. So back here in Studio Web, I'll click the Home icon to go back to my project page. And here I'll click Create New. And for the type of project, I'll choose RPA, or Robotic Process Automation Workflow, because I want to automate the same clicks and actions a person would normally perform in the desktop apps. So let's quickly build a workflow to add a page and a legal notice to my Pages doc. I'll call this Pages Confidentiality Notice, and I'll click the plus button to add an activity to this workflow. The activity I'm looking for first is Use Application Slash Browser, and we can see it appears here. And this activity lets me select which app I want to control with this workflow. So here I'll enable Indicate on Local Machine, which means I can select apps running on this computer. Now I'll click the Indicate on Local Machine button. Again, it's essential that the UiPath Assistant is running for this to work. The first time you do this, it might take several seconds for something to happen, but notice that's minimized the Studio Web browser window, and now I can select from open application windows or any other interface items on my screen. So I'll select the Pages window, and I'll confirm that's what I meant to select in the UiPath Assistant. And now we have our first activity in our workflow, and we can see that UiPath has captured the Pages window as the application context. Next, I want to create a new page in this document. I'll click the plus button inside this application container, which tells UiPath that this action should run in the Pages app. And here I'm going to find the Click activity. This allows me to indicate where in the Pages window to click. So we'll start by clicking the Indicate Target button, which again minimizes my browser window. And now notice as I roll over the Pages window, different interface elements highlight. UiPath uses an AI system called Computer Vision that allows your robot to see all these on-screen elements, which is especially great when there's no traditional programmatic access to these elements. In this case, I want to click the New Page button to automatically add a new page to the document. And I'll click Confirm, which brings us back to our workflow. All right, so after we add a new page, I want to automatically paste in my confidentiality text. So I'll add another activity and search for Set Clipboard. And here I can simply type or paste in the text I want to copy to my computer's clipboard. So I've added the text, This document is confidential and intended solely for internal use. Unauthorized distribution is prohibited. And lastly, we have to tell the robot to paste in that text. So I'll add one more activity of Keyboard Shortcuts. And here we'll add the standard keyboard shortcut for paste, which is Win slash Command and the V key. And I'll click Add Shortcut Key. So at this point, our workflow brings the Pages window to the front, it clicks the New Page button in the Pages toolbar, and it pastes in the text we specified. Let's test it out. Again, I'll click Debug on local machine. And you can see it brings Pages to the front, it moves my mouse up and clicks Add Page. And it adds in that text. So our workflow successfully created a new page and pasted in our text. Being able to program the automation to click on any interface elements really opens up enormous possibilities for your workflows. I also want to point out over here in the log, we see this entry of Healing Agent Configuration. When performing UI automations, your workflow may run into common issues, like for example, if an unexpected pop-up window appears, or the app developer changes the location of the button you want to click. The Healing Agent automatically adapts to common UI changes like these to make sure your automations continue to run seamlessly and are more resilient over time. Now there may also be times when you want to do something a little more complex that doesn't just require clicking interface elements. That's where you can use Screenplay, UiPath's new out-of-the-box agent, which allows you to automate complex UIs with AI just by describing what you want in a prompt. With Screenplay, you can automate any user interface on Mac, whether it's in the browser or on the desktop. Let's create another new project, and again I'll choose RPA Workflow. Now let's say, for example, I want to stay on top of the videos released on the Kevin Stratford YouTube channel. I've opened that page in Google Chrome, and there's the URL for the videos page. So I want to create an automation to automatically open the two most recently published videos in their own tabs whenever I run the workflow. So let's leave this browser window open, and I'm going to switch to the UiPath Assistant. Now to control Chrome, I need to come into Settings, to UiPath Extensions, and here make sure the Google Chrome extension is installed. Similarly, if you want to automate the Safari browser, you'll have to click the link to go to the App Store to install the UiPath Safari extension. In this case, though, we're using Chrome, and I already have the extension installed. So let's switch back to Studio Web in our other browser. So here let's start again by adding the activity of Use Application slash Browser. I'll again switch on the Indicate on Local Machine switch, and then click the Indicate on Local Machine button. I'll give the UiPath Assistant a couple of seconds to start up, and I'll select the Google Chrome window. We'll confirm that. Notice that automatically placed the URL of the page in the URL field. We could also enter it manually, but it was easier to just go to the page first and leave it open so it's automatically entered here for us. Next I'll add another activity, and this time we'll search for and select Screenplay. And again, this uses AI to perform UI tasks on the attached application. So this is such a powerful way to automate. Essentially, instead of having to point at interface elements or explicitly code every step, I just need to write a prompt to tell the AI what I want to have happen. So I'll click in the task field, and I'll type a prompt of, Open the two most recently published videos in their own browser tabs. So I'm not telling it to click on anything, and I don't have to specify a particular thumbnail or indicate anything specific about the page layout. I'm relying on Screenplay to be able to see what's on the page and figure out what the two most recent videos are and open them in their own tabs. Now here in the model menu, you can choose from the different available options that combine UiPath technology with large language models. The default is currently UiPath with Gemini 2.5 Flash. But you may want to experiment with other models if you're not quite getting the results you want. For example, I might choose UiPath with GPT-5. But that's basically it. Let's see if it works. I'm going to close my Chrome window. In fact, let's quit Chrome entirely and see if the automation will open it automatically. And now I'll click Debug on Local Machine. And as you can see, it's opened Google Chrome on its own. It's navigated to the correct URL. You can actually see the thumbnail preview in motion because it's selected that thumbnail. And now it's opened the video in its own tab. And it's moved on to the second most recent video.

[00:13:40] Speaker 2: And there you have it.

[00:13:42] Speaker 1: It's now successfully opened both videos in their own tabs. And I can come back to my computer after letting this automation run and start watching the new videos on the channel. So this would be an incredibly complex automation to build manually, but because we're working with the Screenplay Agent, all we had to do was write a one-sentence prompt to get the results we wanted. So as you've seen, UiPath Studio Web makes it possible to build and run real automations on macOS, from working with local files, to clicking through native apps, to using AI-driven screenplay actions, all from your web browser. And remember, you can sign up and start using UiPath for free at UiPath.com. If you found this video useful and want to see more like it, be sure to give it a like and consider subscribing to the channel. So until next time, I'm Garrick, thanks for watching, and we'll see you soon.