While working on the voice solution at Olympus our team had a lot of debates about whether or not to build a speech-to-text feature in our report application.
I used the OlySense Design System, a React speech-to-text library, and some assistance from Cursor to put together a working prototype to validate the concept.