Build a small Bluetooth device with a push-to-talk control that captures speech, transcribes it locally or through an encrypted service, and sends the resulting text to the active field as standard keyboard input. Preserve every recording and transcript until insertion succeeds, support spoken editing commands and personal name corrections, and provide an optional setup and training service for users with disabilities.

Voice transcription quality is now strong enough for practical daily use, but mobile operating-system restrictions still make third-party keyboard implementations unreliable. Repeated willingness to pay, paired with complaints about high subscriptions and accessibility barriers, creates room for a dependable hardware-enabled alternative with transparent pricing.