Write ups at speaking speed
Dictate letters, reports and notes in the record system you already have open.
For healthcare
Notes that never have to leave the room
Clinical language is specific and patient information does not belong in someone else's cloud by accident. Mr.Speech has a mode where speech recognition runs entirely on your own machine, and a dictionary that learns your terminology.
- On-device speech recognition
- No account needed for local mode
- Personal terminology dictionary
- You choose per task
For healthcare
How it fits the day
01
A local mode that is actually usable
The on-device engine handles 40 languages with real punctuation and runs about 7 to 9 times faster than real time on an ordinary processor. It is fast enough to use all day, which is what makes choosing it realistic.
02
It learns clinical terminology
Drug names, procedures, abbreviations and local conventions go into a personal dictionary rather than being mangled the same way every time. Corrections you make are offered back as one click suggestions.
03
It types into the system you already use
There is no separate window and no copy and paste step. Text lands in whatever field has focus, including a browser-based record system.
04
The field-reading feature is optional
To place text correctly, Mr.Speech can read the field you are typing into. That is a setting you control, and it never touches password fields.
05
No emotion inference, ever
There is no sentiment scoring, no mood detection and no stress metric anywhere in the product. EU law prohibits inferring emotions from biometric data at work, and we did not want to be near that line in the first place.
For healthcareDocumentation Teaching Control
Where it earns its place
Lectures and case discussions
Capture a room with a single microphone, with live subtitles for the audience if you are presenting.
Nothing uploaded by accident
Local mode needs no account and no internet, so there is no path for audio to leave the machine.
Questions
Questions
Do you claim HIPAA or any medical certification?
No. We make no compliance claims and offer no badge we cannot evidence. What we do is tell you precisely where processing happens, so your organisation can assess it. Start with our security page.
Does the local mode really work without internet?
Yes. Speech recognition, semantic search and meeting transcription all have full on-device implementations. No account, no connection.
Can it analyse how a patient sounds?
No, and it never will. There is no sentiment or emotion analysis in the product. Our EU AI Act page explains the reasoning.
