FAQ
The questions we get most. If yours is not here, write to us at hello@voiceflo.io.
Does this work in a noisy plant?
Yes, with a caveat. We use Whisper-class transcription which handles loud rooms better than older speech engines, but the closer the mic is to the speaker, the better the read. Bluetooth headsets and shoulder mics work great. We also let you replay the original audio on any record, so if the AI mishears, your team can verify in seconds.
What happens if a worker has gloves on?
They press once and talk. The big mic stays on screen on every page. There is no menu to navigate, no keyboard. If the gloves are wet, that is fine too, the mic is a single oversized target.
Do my workers need to learn a syntax?
No. They just talk. The orchestrator figures out the intent. "Open the temperature check" works. "Pull up that temp form for cooler 3" also works. If it can not tell, it asks a short question back.
How do you handle data we can not get wrong, like CCP logs?
We never auto-submit forms that matter for compliance. The voice fills the field, your operator confirms with a tap. Out-of-spec readings flag the supervisor in the same second they get logged.
Can supervisors and admins build workflows by voice too?
Yes. An admin can say "Create a new workflow called pallet rotation" and Voiceflo opens the builder pre-named. They can keep going by voice ("add a number field called pallet count") or finish on a desktop. Workers stay focused on filling forms.
Where does the data live?
In your Supabase project, with row-level security per enterprise. We never see customer data outside of the function call that runs the AI. You can self-host the whole stack if your compliance team requires it.
How does pricing work?
Per active worker per month, with a flat rate for the back office. No per-transcription metering, no overage surprises. Get in touch for a number.
Do you integrate with our WMS or ERP?
We push records out through webhooks and a REST API. We have done direct integrations with NetSuite and SAP for pilot customers. If you have a system we have not seen, send us the docs and we will scope it.
What about offline?
The mobile app records, queues the audio locally, and uploads when the device is back online. The user gets a clear "queued" indicator so nothing feels lost.
Is the AI making things up?
The AI does two things: transcribe what was said, and pick a structured action. It does not invent values. Every record stores the raw transcript next to the parsed value, so your team can audit every captured number against the audio.
Pilot Voiceflo on one line, one shift.
We set it up with you, train your supervisors, and you decide if it stays after two weeks.