Built for every patient.
Including ones who don't speak with their voice.
If you are Deaf, Hard-of-Hearing, use a sign language, or rely on a pictogram board, P.AI treats your communication preference as a clinical signal, not an afterthought.
Tools you can use right now
Every page has a violet accessibility button in the bottom-right corner. Tap it to open these tools. Your settings are remembered.
Larger text
Two sizes available: Large (×1.15) and Larger (×1.30). Affects every page until you turn it off.
High contrast
Yellow-on-black, AAA-friendly. Designed for low vision, AMD, diabetic retinopathy, light sensitivity.
Read aloud
Browser-native text-to-speech reads the current page. No data leaves your device. Works in Chrome, Edge, Safari, and most modern browsers.
Screen-reader friendly
Live regions (aria-live), descriptive labels, semantic HTML. Compatible with NVDA, JAWS, VoiceOver, TalkBack.
For Deaf / Hard-of-Hearing — what we actually do
Sign language as first-class option
Choose ASL · ISL · BSL · LSF · Auslan in intake. Encounter flagged for clinic.
Interpreter request — in-person or VRI
One toggle. Choose modality: in-person, VRI, phone, family (not preferred for clinical).
AAC / pictogram users supported
If you use an AAC device, mark it in intake. Voice-only flows are disabled.
Section 1557 / ADA
Federal law requires the clinic to provide a qualified interpreter at no cost to you.
What we are NOT pretending to do
Honesty over hype.
- We do not read sign language via camera. That's a multi-decade research problem, not a feature.
- We do not auto-translate to plain language using AI. Your intake captures what you said in your own words. Your visit summary uses standard clinical language because your doctor needs it that way.
- We do not lip-read or use AI gaze tracking. Voice intake (Sarvam STT) is silently disabled when you choose Sign Language or AAC.
- Read-aloud uses your browser, not the cloud. No data leaves your device when you tap "Read aloud."