A different way to write
Why this product exists, what it refuses to do, and the concrete choices guiding its design.
The project
Dictation has been around for a long time, but speaking loud enough to be understood exposes the entire conversation. In a shared office, a train or a waiting room, this is often enough to give up.
Speakey starts from the opposite gesture. We want you to be able to hold a key, speak quietly and find the text at the cursor in the app that is already open.
Dictating should not require the entire room to be heard.
The speech-recognition engine and application are in development. We want to operate the future service in Europe and publish clear rules on data use and retention before the first pilots. The intended choices are presented under Your data.
The project will move forward in stages: first validate quiet-speech recognition, then integrate it into an application, open pilots and finally offer a downloadable release. Features and platforms will be announced as they are confirmed.
How we move forward
Prove before claiming
Quality, latency and quiet speech will be measured with an understandable protocol. An intention is not presented as a result.
Start from real work
Future pilots must show where voice truly helps, where it gets in the way and which corrections remain necessary.
Make choices understandable
Processing, retention and providers must be explained before anyone entrusts their voice to the product.
Want to dig deeper or challenge us on privacy? Write to us.