Trij — On-device AI medical triage for community health workers

Trij is an offline-first progressive web app that brings AI-assisted medical triage to community health workers in remote areas. Powered by Google DeepMind's Gemma 4 models — entirely on-device.

Built for the Gemma 4 Good Hackathon (Kaggle + Google DeepMind, $200K prize pool).
Track: Health & Sciences / Global Resilience.

---

Quick demo

Take a photo → AI assesses → Get urgency + recommendation → Save offline → Auto-sync

No internet needed. No patient data leaves the device.

---

Features

• 📸 Wound & rash triage — Snap a photo, Gemma 4 analyzes it on-device
• 📄 Medical document scanner — OCR + analysis of lab reports, prescriptions
• 🎤 Voice-guided assessments — Speak in your language, get spoken responses
• 📋 Patient records — Create, view, and track patients offline
• 🏷️ Urgency triage — Green (routine), Yellow (soon), Red (urgent)
• 📎 Referral PDFs — Auto-generated referral slips for clinic handoff
• 🔄 Offline-first sync — All data stored locally, syncs when online
• 🌍 Multilingual — English, Spanish, French, Swahili, Hindi, Arabic, Portuguese
• 🔒 Privacy-first — All AI runs on-device. No cloud AI API.

---

Tech stack

Layer	Technology
Frontend	Vite + TanStack Start + React 19 + TypeScript
AI	WebLLM (WebGPU) + Ollama bridge + Demo mode
Styling	Tailwind CSS v4 + shadcn/ui
Offline	Dexie.js (IndexedDB) + background sync
Backend	Supabase (Auth, Postgres, Storage, RLS)
Voice	Web Speech API (7 languages)
PWA	vite-plugin-pwa (installable on Android/iOS)

---

Getting started

Prerequisites

• Node.js 22+
• Bun (recommended) or npm
• A Supabase account (free tier)
• (Optional) Ollama for local Gemma 4 inference

1. Clone and install

git clone https://github.com/Mosss-OS/trij.git
cd trij
bun install

2. Set up Supabase

1. Create a project at supabase.com
2. Copy .env.example to .env and fill in your Supabase credentials
3. Run migrations:

npx supabase db push

3. Set up Gemma 4 (choose one)

Option	Setup	Notes
Demo mode	Nothing to do	App works immediately with mock data
Ollama	ollama pull gemma4	Best real-model experience on laptop
WebLLM	Requires Chrome + WebGPU	Loads ~1.5GB model on first triage

For Ollama:

ollama pull gemma4
# Or download from Kaggle: scripts/download-gemma4.sh

4. Run

bun run dev

Open http://localhost:5173 — sign up, and you're ready.

Alternative: Docker (full stack)

cp .env.docker .env
docker compose up

This starts the app (hot-reload), Ollama with Gemma 4, and a local Supabase stack.
Open http://localhost:5173.

---

Project structure

src/
├── routes/          # TanStack file routes
│   ├── index.tsx    # Login / sign-up
│   ├── _app.tsx     # Auth layout
│   ├── _app.dashboard.tsx
│   ├── _app.triage.tsx        # Camera → AI → result
│   ├── _app.document.tsx      # Document scanner
│   ├── _app.patients.index.tsx
│   ├── _app.patients.$patientId.tsx
│   ├── _app.supervisor.tsx
│   └── _app.settings.tsx
├── components/      # Reusable UI components
├── lib/             # Core logic
│   ├── gemma.ts         # WebLLM + Ollama + demo engines
│   ├── gemma-prompt.ts  # System prompts for Gemma 4
│   ├── db.ts            # Dexie IndexedDB schema
│   ├── sync.ts          # Background sync engine
│   ├── voice.ts         # Speech I/O
│   └── referral.ts      # PDF generation
├── hooks/
│   ├── useGemma.ts      # Model lifecycle hook
│   └── useOnlineStatus.ts
├── stores/          # Zustand state
└── types/

---

Hackathon submission

• Technical writeup: TECHNICAL_WRITEUP.md
• SRS: SRS.md
• Development prompt: DEVELOPMENT_PROMPT.md
• Model setup: scripts/download-gemma4.sh

---

Contributing

See CONTRIBUTING.md for development setup, code conventions, and pull request process.

---

License

Apache 2.0 — see LICENSE.

---

Built with ❤️ for the Gemma 4 Good Hackathon.