AI Voice Assistant for Windows

An assistant that actually presses the buttons

Most desktop assistants either answer questions or automate tasks. This one does both. Ask it
something and you get a spoken answer; tell it to do something and it does it — opens File Explorer
at a specific folder, launches an application, runs a web search and brings up the results, finds a
document you half-remember, mutes your audio or takes a screenshot.

How a request flows

1. Say the wake word, or press your hotkey.
2. Speech is transcribed and turned into an intent.
3. The matching action runs on your PC, and anything worth saying out loud is spoken back.

Because intent matching is natural language rather than a fixed phrase list, "pop open my downloads
folder" and "show me downloads" both land in the same place.

Things people ask it every day

• "Open File Explorer in Documents"
• "Search the web for the best mechanical keyboards in 2026"
• "Find the invoice PDF I saved last week"
• "Launch Visual Studio Code"
• "Set volume to thirty percent"
• "Take a screenshot and put it on my desktop"
• "What's the weather tomorrow?"
• "Start a twenty-five minute timer"

Privacy by design

The wake word runs entirely on your machine, so the microphone is not streaming anywhere while you
work. Only the request you actually speak is processed, you can see a full history of what was heard
and what was done, and you can switch to push-to-talk if you would rather not use a wake word at
all.

Getting started

1. Download and extract the ZIP.
2. Launch the app and pick your microphone.
3. Choose a wake word or a push-to-talk hotkey.
4. Say "open File Explorer" to confirm everything works.