# Computer Controller Most AI tools stop at the browser. The moment work lives in Photoshop, Excel, Logic, Final Cut, or a native app with no API, they're stuck. Computer Controller is what lets an employee step outside the browser and operate your actual desktop, real mouse, real keyboard, real screen, through the companion app running on your machine. Beyond the browser, your employee controls desktop applications. It opens apps, navigates menus, clicks interface elements, moves files between folders, takes screenshots, and performs tasks that normally require a human sitting at the keyboard. If you can do it with a mouse and keyboard, your employee can do it too. Consider a workflow that lives partially in a desktop app: open the company's accounting software, export the monthly report as a PDF, move it to a specific folder, rename it with today's date, and email it to the finance team. Your employee handles the entire sequence. No API needed for the accounting software. It just operates the application the way you would. Desktop control works across macOS, Windows, and Linux. Every action is logged and visible in the activity feed, so you can review exactly what your employee did on your machine. Combined with the approval gateway, you maintain full control over which desktop actions proceed and which require your sign-off. The desktop companion app is a lightweight application you install on your computer. It bridges your AI employee to your local machine, enabling both browser automation and desktop application control. Without it, your employee works through connected cloud tools. With it, your employee can also operate anything on your screen. ## AI Agents That Operate Your Desktop Applications Desktop Application Control extends your AI employee beyond the browser and into native apps: spreadsheet software, creative tools, data analysis platforms, internal applications with no web version. If it runs on your machine, the agent can operate it. The agent interacts with desktop apps the way a human would: moving the mouse, clicking menus, entering text, reading screen content. It can open a file in Excel, run a calculation, copy the result, and paste it into a report, all from one instruction. No macros. No scripts. Just an autonomous agent doing the work. ## File Operations and System-Level Control Beyond app control, Desktop Automation includes file system operations: move files between folders, rename batches, compress archives, organize downloads, copy outputs from one location to another. Tedious file management tasks that take humans twenty minutes take the agent twenty seconds. Screenshots are captured automatically during automation runs, giving you a visual record of every step. The agent can also take on-demand screenshots to verify a state, check for errors, or confirm that a form was submitted correctly. Full observability into what the agent is doing on your machine. ## Runs on macOS, Windows, and Linux Desktop Automation works across all major operating systems. The Desktop Companion App establishes the local bridge and handles OS-specific interactions transparently. Your AI agent does not need to know which OS it is on, it just operates the apps you have. Cross-platform support matters for teams with mixed environments. A marketing team on Mac and a data team on Windows can both deploy desktop-capable AI employees without platform-specific configuration. One workforce, any machine. ## A Lightweight Bridge Between Your AI Employee and Your Machine The Desktop Companion App is a small, always-running background application that gives your AI employee access to your local machine. Install it once, and the agent gains the ability to control desktop apps, manage files, and take screenshots, all coordinated from the Sistava cloud platform. The app is not a heavy client. It does not run the agent or the model. It is purely a bridge: it receives instructions from the cloud, executes local actions, and reports back. Your machine stays in control. The agent stays in the cloud. ## Desktop Automation vs Traditional RPA Traditional RPA tools automate the desktop by recording fixed coordinates and brittle selectors. The moment a button moves or an app ships an update, the recording breaks and someone has to rebuild it. That fragility is why so many RPA projects stall, and why they usually need a dedicated team and an expensive license to maintain. An AI employee operates the desktop the way a person does. It reads the screen, finds the element it needs, and adapts when the interface changes, so there is no script to re-record after every update. You describe the outcome in plain language and the agent works out the steps. That is what finally makes desktop automation practical for a small team that has no RPA budget and no time to babysit flowcharts. ## What Your AI Employee Can Automate on the Desktop Accounting and finance tools such as QuickBooks, Xero, and legacy ERP systems: enter transactions, reconcile accounts, run reports, and export statements. Spreadsheet work in Excel and Numbers: open a workbook, run a calculation, clean a dataset, and paste the result into a report. Creative and design software: open project files, apply consistent formatting, and export assets in batches. Internal and proprietary applications with no API or web version: the agent operates the windows, menus, and forms directly, so the tools that were impossible to automate become part of your workflow like any other. ## Computer Use AI, Ready for Real Work Computer use, AI that operates a real machine with mouse and keyboard, is one of the most demoed capabilities in the industry. The hard part is making it dependable enough for a business to lean on. That is what the surrounding platform is for: approval gates in front of consequential actions, screenshots and logs for every run, and schedules that turn one-off tricks into recurring work. It also compounds. The employee that operates your desktop is the same one that knows your business, remembers your conventions, and works your other channels, so desktop control is not an isolated gadget. It is one more place your workforce gets things done. Most public demos of computer-use AI are one-off: a fresh session with no memory of your conventions, starting over every time. This is not a demo. It shares the same memory the employee uses on every other channel, so if you have corrected how a report should be formatted or where files belong, the next desktop run applies that without being told again. ## How It Works **Your AI employees control native desktop applications directly, handling tasks in software that has no API and no browser interface.** Install the Sista companion app on macOS, Windows, or Linux and your AI agents gain access to your local desktop environment. The agent can open applications, read screens, navigate menus, type text, move files, take screenshots, and interact with any GUI element. This covers accounting software, design tools, internal enterprise systems, and anything else that runs as a native app. Desktop automation works alongside browser automation, so a single workflow can span multiple environments. The agent might pull data from a web dashboard, open a desktop application to process it, and save the output to a local folder, all in one unattended run. Approval gates apply here too: the agent requests permission before any action that modifies files or submits data in a native app. ## Use Cases ### Finance team automates legacy accounting software An AI agent opens the desktop accounting app, enters transactions, runs reports, and exports the output, replacing hours of manual data entry. ### Design team batch-processes files in local software The AI employee opens design files in desktop tools, applies consistent formatting, exports assets, and organizes the output folder. ### Operations team works with apps that have no API Any desktop application becomes automatable. The AI agent clicks, types, and navigates just like a human operator would. ### IT team automates system configuration tasks An AI agent opens system settings, applies configurations, and verifies changes across machines, reducing manual IT workload. ### Support team pulls records from a legacy desktop CRM The AI agent opens the on-premise CRM, looks up a customer, copies the account history, and drops it into the ticket, with no integration to build. ### Analyst compiles a report across desktop and web The agent pulls figures from a web dashboard, processes them in a desktop spreadsheet, and saves the finished report to a local folder in one unattended run. ### Team member delegates a desktop task by phone while away from their desk Call the AI employee, describe the file cleanup or report export needed, and it completes the task on the companion machine using the same steps it would take from a typed instruction. ## Comparison | Before | After | |---|---| | Legacy desktop software cannot be automated via API. | The AI agent controls any desktop app, old or new, with or without an API. | | Manual data entry into desktop tools costs hours per week. | The AI employee handles data entry automatically, with no errors. | | Desktop workflows are excluded from automation pipelines. | Desktop and web steps combine in one seamless agent workflow. | | Automating desktop apps requires expensive RPA tooling. | Desktop automation is built in, no extra software needed. | ## FAQ ### Does desktop automation require installing software on my machine? Yes. The Desktop Companion App must be installed on the machine the agent will control. It is a lightweight bridge that gives the agent access to local apps and the file system. The Sistava web platform manages the agent itself. ### What operating systems are supported? Desktop Automation supports macOS, Windows, and Linux. The Desktop Companion App handles OS-specific interactions so the AI employee works consistently across all three. ### Is desktop automation safe for sensitive work? The agent only operates apps and files within the scope you define. Approval gates can be configured for sensitive actions like deleting files or submitting forms. All actions are logged with screenshots for full auditability. ### Can the agent automate apps that have no API or browser interface? Yes. Desktop automation works with any app that runs natively on your machine, including legacy tools, proprietary internal software, and apps with no web or API access. ### Does the Desktop Companion App run the AI model locally? No. The app is a lightweight bridge only. The AI model and all agent logic run in the Sistava cloud. The desktop app handles local execution when the agent needs to interact with your machine. ### How is this different from RPA tools like UiPath or Power Automate? Traditional RPA records fixed clicks and selectors that break when an app updates its layout. Your AI employee reads the screen and adapts the way a person does, so there are no brittle scripts to maintain. You also brief it in plain language instead of building flowcharts, and it can combine desktop and browser steps in a single workflow. ### Can the agent run desktop automations unattended or on a schedule? Yes. Once the companion app is installed and running, the agent can execute desktop workflows on a schedule or a trigger with no one at the keyboard. Approval gates still apply to any sensitive step you choose to gate, and every run is logged with screenshots so you can review it after the fact. ### What happens if the agent hits an unexpected screen or an error? Because the agent reads the screen at each step rather than following fixed coordinates, a moved button or a new dialog does not break it. If it cannot proceed, it captures a screenshot, reports the state in the activity feed, and can pause for your input instead of failing silently. ### Does it work with specific apps like Excel, QuickBooks, or SAP? Yes. If the application runs on macOS, Windows, or Linux and you can operate it with a mouse and keyboard, the agent can operate it too. That includes spreadsheets, accounting software, ERP systems, and proprietary internal tools that have no API. ### What data leaves my machine when the companion app runs? The companion app receives instructions from the Sistava cloud and executes them locally. The AI model runs in the cloud, the actions run on your machine, and you control which apps and folders are in scope. Every action is logged so you can review exactly what happened. ### Can I watch the agent work in real time? Yes. Actions happen visibly on the machine, and every run is logged with screenshots in the activity feed, so you can watch live or review afterward. ### Do I need the companion app on every machine the agent controls? Yes, one install per machine you want it to operate. The agent itself stays in the cloud, so the same employee can work across any machine that has the companion running. ### Can I ask for a desktop automation task by voice instead of typing it? Yes. Voice is one of the channels your AI employee already supports, so you can call and describe the task the same way you would type it, and it carries out the same desktop steps once the companion app is running. ### Does the agent remember how I like files organized or reports formatted between runs? Yes. The same memory system that carries context across conversations applies here, so once you have corrected how something should be done, the employee applies that going forward without you repeating the instruction. > We run a proprietary desktop tool with no integrations. The AI agent now operates it directly, which saves my team about eight hours of manual entry per week. > > Aisha O., Data Operations Lead ยท mid-size company ## Where Computer Controller fits Computer Controller is part of How they use your computer. Your AI agent operates your browser and desktop apps the way a human would. Navigate web apps, fill forms, click buttons, move files, and complete multi-step workflows across any application. Install the lightweight companion app and your employee gains access to your entire machine. - [How they use your computer](/en/features/automation): Control any browser or desktop app. ## Read the guide - [Guide: Computer Controller](/en/guide/capabilities/computer) ## More in Desktop Control - [Browser Automation](/en/features/automation/browser_automation): Most automation tools can read a public web page and stop there. The moment a task needs a real login, a multi-step form, or clicking through an actual workflow, they fall over. Browser Automation gives an employee a real browser on your own machine, driven through the desktop companion app, so it can navigate, click, type, and pull data out of sites the same way you would, including the ones that live behind a sign-in. - [Proactive Employer Messaging](/en/features/automation/proactive_employer_messaging): Your AI employee does not wait for you to check in. When it finishes a task, hits a blocker, or needs a decision from you, it sends the update itself, straight into the same chat thread you already use with that employee. No separate inbox, no polling a dashboard to see if something happened. - [Automatic File Organization](/en/features/automation/file_organizer): Your AI employee sorts, renames, and files documents on your own computer through the Sistava Desktop Companion app: reading what is in a folder, working out what each file is, and moving it where it belongs using the naming convention and folder structure you describe in plain language. File access stays inside your home folder, and every move, rename, or write is recorded in the activity feed so you can see exactly what changed. - [Hands Free Data Entry](/en/features/automation/data_entry_automation): Your AI employee reads records from a spreadsheet, export, email, or PDF and types them into any web or desktop application, including tools with no API. Every field is checked against the source before submission, and anything incomplete or ambiguous is set aside in an exception report instead of entered wrong. Recurring entry jobs run on a schedule with the approval checkpoints you choose. ## Explore - [Every feature](/en/features) - [Hire an AI employee](/en/market) - [Pricing](/en/pricing)