Research & Constraints
I started by speaking informally with two physicians about how they document today. A few patterns stood out:
- They don’t want to touch the screen more than necessary
- Real-time transcription is useful, but only if it’s reliable
- Most errors happen when the tool misunderstands a word — and correction workflows are tedious
From there, I reviewed existing tools used in clinics and noted key gaps:
- Hard to access – installation is usually handled by IT; setup is slow and rigid
- Outdated UX – interfaces feel clunky and are not designed for modern workflows
- Lack of intelligence – most tools just directly transcribe speech without smoothing, structure, or prioritizing clarity
- Poor mobile support – limited or no mobile-first options, even though many doctors rely on tablets and phones throughout the day
Designing the Flow
I mapped a 3-step user journey: Dictate → Review → Edit & Export
I sketched wireframes for both mobile and desktop, prioritizing:
- One-tap recording
- Visual feedback during speech
- Simple correction without typing (e.g. tap-to-edit)
Visual Design
The UI is clean, muted, and distraction-free. I used subtle colors for feedback states (e.g. grey for captured text, yellow highlights for unsure words). Typography is large and readable, especially on mobile.