Manual Privacy
Where your voice and text go
What leaves your PC when you dictate, what never goes anywhere, and the few things that can change that.
With Neato Echo's default setup, nothing leaves your PC. Audio is transcribed on your machine, the transcript is stored on your machine, and summaries, cleanup and the agent run on a language model on your machine. Neato Ventures has no servers in the path and never sees your audio or text.
Three things can change that, and each one is something you turn on yourself.
The engine is set three times, not once
Dictation, note recording and audio upload each have their own speech-to-text engine setting. They're independent, so it's worth checking all three if you ever change one.
Open Settings, choose Speech-to-Text, and you'll see three tabs:
| Tab | Covers |
|---|---|
| Dictation | Press-to-talk dictation |
| Note Recording | Meeting and note recordings |
| Audio Upload | Files and URLs you upload |
All three default to a local engine (NVIDIA Parakeet). Changing one does nothing to the other two.
What each setup does with your audio
| Setup | Where the audio goes | What Neato Ventures keeps |
|---|---|---|
| Local (default) | Nowhere — transcribed on your PC | Nothing |
| Cloud provider, your own key | Straight to the provider whose key you entered | Nothing |
If you add a key under Settings → Cloud Providers and point an engine or the language model at it, the request goes from your PC directly to that provider. We don't proxy it and we don't see it; that provider's terms and privacy policy govern what happens next.
What travels with a cloud provider request
If you've chosen a cloud provider for transcription, each request carries:
- the audio itself
- the model you're transcribing with
- your transcription language, if you've set one rather than using auto-detect
- your custom dictionary terms, if you use one — they're sent as a prompt so the model spells names and jargon correctly
That last one surprises people, so it's worth stating plainly: if you've added client names or internal project names to your dictionary, those words go with every cloud transcription request. If that's not acceptable for your setup, keep the dictionary empty or stay on a local engine.
If you've chosen a cloud provider for the language model, the text of your dictation, note or question goes to that provider, along with any notes the agent reads to answer you.
Screen context is the one case where an image is sent. If you turn on Share screen context — it's off unless you do — then pressing the voice agent hotkey also sends a screenshot of the display your cursor is on to whichever model the agent is pointed at. With a local model that stays on your PC; with a cloud provider it goes to them. It's used for that single request and nothing else: never written to disk, never added to your notes or transcription history, and never included in logs. Ordinary dictation never sends one.
Calendars
If you connect a Google or Microsoft calendar in Settings → Integrations, Neato Echo reads your upcoming events from that provider so it can remind you about meetings and link notes to them. Access is read-only, the events are cached in the local database, and nothing is written back. Disconnecting stops the sync.
Updates
At launch the app asks github.com whether a newer release exists. That request carries the app version and nothing about you or your content.
Web search
The chat agent can search the web only if you've set a cloud provider key that supports it. With no key, there's no web search and no search queries leave your PC.
Usage analytics
There are none. Neato Echo sends nothing about your usage to anyone.
Related
Still need help?
Write to [email protected] with your Windows version, your Neato Echo version (Settings → System), and what you were doing. A person reads every message.