Accessibility & Language Support

Issue Date: 2025-12-09 Document Version: v1.1.0

Version Log

v1.1.0 (2025-12-09): Updated Issue Date, Document Version, and clarified accessibility/language sections; no material changes to commitments or scope. v1.0.0 (2025-12-05): Initial publication covering accessibility features, WCAG compliance, and multilingual support guidance.

Purpose

This document outlines PepperChat's commitment to accessibility and provides guidance on language support for users and clinicians.

Accessibility Statement

We design PepperChat in alignment with WCAG 2.2 Level AA guidance and human‑centered design practices (ISO 9241‑210). Information presentation follows ISO 9241‑112 principles for clarity.

What we support today:

Screen reader compatibility: semantic HTML, labeled form controls, ARIA where appropriate, and live‑region announcements for dynamic content

Keyboard‑only navigation: all interactive elements are reachable with the keyboard; focus order is logical; visible focus styles are provided

Text scaling: supports browser zoom and OS‑level text size adjustments up to 200% without loss of content or functionality

Contrast: color palettes target a minimum 4.5:1 for body text and 3:1 for large text and UI elements

Structure and headings: consistent page structure, landmarks, headings, and labels to aid navigation

Error handling: inline messages are associated to inputs and exposed to assistive technologies

Standards:

WCAG 2.2 Level AA alignment

Automated checks (axe/Lighthouse) and manual reviews with screen readers

Contact and feedback:

If you encounter an accessibility issue or need content in an alternative format, contact support@pepperchat.ai or visit the Support page in the app.

Roadmap:

Improve skip‑to‑content and landmark coverage

Continue contrast and focus‑ring refinements

Expand end‑to‑end assistive technology testing

Language & Multilingual Support

How to use today (UI + AI):

Translate the interface with your browser

Chrome/Edge: right‑click → "Translate to …" or use the translate icon in the address bar; you can set "Always translate" for our site

Safari (macOS / iOS): tap the aA icon → Translate → choose your language

Ask the AI to respond in your language

Add a simple instruction to your prompt, for example:

"Summarise the session in Spanish (LatAm). Keep a formal clinical tone."

"Generate goals and interventions in French (Canada), ICD‑aligned language."

"Write the progress note in Portuguese (Brazil). Do not mix languages."

Tip: specify the locale (e.g., es‑MX, fr‑CA, pt‑BR) for region‑appropriate spelling and phrasing.

Summary:

The UI and validated AI workflows are currently in English. The underlying model supports many languages and can draft in a requested language while preserving clinical structure.

Supported today:

UI language: English

AI drafting/summarisation: optimised for English input/output; can generate in a specified language upon request

Dialect handling: English variants (US, UK, AU, CA)

Roadmap:

Add first‑class multilingual support for transcription and AI outputs (targeting Spanish, Portuguese, French). When enabled, PepperChat will detect preferences and return outputs in the requested language, using language‑specific models where available.

How AI adapts to language:

User preference: an upcoming setting will let you choose a default output language

On‑the‑fly: you can request a specific language in your prompt

Clinical clarity: outputs maintain clinical structure and ICD‑aligned terminology appropriate to the target language where applicable

Compliance note:

As with any documentation tool, clinicians should review final phrasing for local payer/insurer requirements. We will continue clinical review as we expand language support.

User Testing Across Languages

We have not yet conducted structured user testing in languages other than English. As we expand multilingual support, we will incorporate testing with clinicians and users who speak Spanish, Portuguese, French, and additional languages. This will include testing with non-native English speakers and individuals with varied accents or dialects to evaluate clarity, usability, and AI output consistency. Findings from this testing will inform future refinements before full multilingual capabilities are released.

Document Control

Owner: Product & Engineering

Review Cadence: At least annually, and after material accessibility or language feature changes

Next Scheduled Review: 2026-01-31