Windows applications and files
Interact with supported application interfaces, keyboards, mouse actions, dialogs, folders, documents, and local data.
Automate tasks that move between Windows software, browsers, files, spreadsheets, APIs, and connected Android devices. Let AI draft the visual flow, then run the approved commands on your machine.
Real tasks rarely stay inside one website. A record may begin in a spreadsheet, require a browser lookup, continue in desktop software, and finish with a file, API request, or mobile action.
Interact with supported application interfaces, keyboards, mouse actions, dialogs, folders, documents, and local data.
Capture web elements, read and enter data, follow pagination, and coordinate supported parallel browser sessions.
Use supported mobile element and device commands, then pass results directly into desktop or browser steps.
Desktop and browser commands execute on the Windows machine where Runavelo is installed. This is useful for software, files, devices, and authenticated sessions that are not exposed as public cloud APIs.
Local does not mean that every workflow is offline. A workflow can intentionally call an API, use an online recognition service, publish an application, or send a prompt to your selected AI provider. Those boundaries stay visible in the command sequence and configuration.
For sensitive tasks, review which commands make network calls, limit the data passed to external services, protect credentials, and test the workflow with representative non-production data first.
Describe the outcome, applications, inputs, outputs, and important constraints. AI can assemble the initial workflow from documented Runavelo commands. The result opens in the visual editor, where every command and parameter can be changed.
You can continue the conversation to add a branch or improve error handling, or make a small manual change when that is faster. The assistant can also explain how to drag and configure commands yourself.
Ordinary workflow commands execute deterministically. That means the same defined logic is reused without spending model tokens on every repetition. AI is brought back when a person asks for a revision, teaching, or troubleshooting—or when the workflow deliberately includes a model-powered command.
Use variables, conditions, loops, subworkflows, retries, and logs to make the behavior explicit and testable.
A workflow can combine supported Windows applications, browsers, files, folders, spreadsheets, APIs, code, and connected Android devices. Capabilities vary by edition.
Desktop and browser runtime commands execute on the Windows machine. Data leaves the machine only when the workflow uses an external API, cloud service, or configured AI provider.
No. Personal accounts use email sign-in and do not require a Microsoft organizational tenant or administrator configuration.
Yes. Advanced workflows can include supported Python, JavaScript, SQL, API, and custom-skill capabilities where appropriate.
Start with the Personal Free edition, view personal access plans, or contact us about enterprise workflows and deployment.