Hold, whisper, read
This simulation illustrates the interaction we want to make possible: hold a key, speak quietly and find the text where the cursor is. The application and its speech-recognition engine are in development.
Explore the intended interaction
Hello Camille, it is noted for Thursday at two p.m. I will send you the report the day before, with the two quotes.
Click the microphone to dictate, move it if needed · in the app, you hold down a key
This is not a technical demo. The text is pre-written, and this page does not capture, store or send any sound.
Before you begin
Three things to know, and you know it all.
- 01
One key to dictate
The goal is for the microphone to be active only while the key is held down and to switch off as soon as it is released.
- 02
Voice corrections
We will explore commands to shorten, rephrase or continue text in another language without returning to the keyboard.
“Shorter”“More formal”“Continue in French” - 03
One correction that remains useful
Correct the spelling of a client name and Speakey can reuse that spelling in later dictations.
What this simulation can demonstrate
The sequence of actions
It shows the intended experience: activate input, see text appear and continue working in the same place.
Not technical performance
It does not measure accuracy, latency or quiet-speech recognition. Those results must come from the future engine and a published protocol.
Not a list of promised features
The commands and personalisation shown are product directions. Their order, form and inclusion in the first release remain to be validated.
Would you like to follow development and hear when the first pilots open?