We kept seeing the same scene in medical practices, offices and technical teams: a good part of the workday turns into record-keeping. The appointment is over, and someone has to turn what they already know into text inside a system. The work is done; what’s left is typing it up, competing with the next appointment.
The obvious way out would be dictation, and almost everyone had already tried it. What got in the way was always the same detail: the dictation services built into the operating system or the browser send the audio to another company’s server. A patient note, a deposition, a clause, an account number: it all leaves the machine before it becomes text. In a regulated industry, that ends the conversation.
There was one more detail the sales material for these services doesn’t mention: no internet, no dictation. The tool fails exactly where it would be most useful: on a plane, in an area with a weak signal, in a practice with an unstable network. It was the sum of these three things that convinced us to build instead of recommending something off the shelf.
