vinogola/dawa-scout

AI-powered triage dashboard for Community Health Workers in rural Kenya, aligned to Kenya IDSR 2023 thresholds — offline-ready PWA. # Dawa Scout **Live:** dawa-scout.vercel.app AI-powered triage dashboard for Community Health Workers (CHWs) in rural Kenya. Dawa Scout checks patient symptoms against Kenya IDSR 2023 / IMCI thresholds in real time, flags cases that need escalation, and gives supervisors a single queue to review referrals — designed to keep working offline in low-connectivity areas. > Built as a learning project (PLP). Data shown in the demo is synthetic — no real patient > records are used. ## Features - **CHW mobile interface** (`/chw`) — a triage form optimized for a 360px screen, matched against IDSR/IMCI/ANC thresholds as symptoms are entered. - **Supervisor dashboard** (`/supervisor`) — a queue of visits by risk level (HIGH/MEDIUM/LOW), with guard-triggered escalations surfaced first. - **Offline-ready (PWA)** — installable to a home screen, with a service worker caching the app shell so it keeps working without a connection. ## Tech stack React + TypeScript, Vite, Tailwind CSS, Radix UI primitives, React Router, `vite-plugin-pwa`. ## Getting started ```bash npm install npm run dev ``` Other scripts: `npm run build` (type-check + production build), `npm run lint`, `npm run preview`. ## Project structure ``` src/ views/ # LandingPage, CHWView, SupervisorView components/ # shared UI (ViewToggle, DarkModeToggle, InstallBanner, ui/) context/ # ThemeContext data/ # mockData.ts — synthetic demo visits lib/ # utils ``` ## Status Prototype stage — triage data is mocked locally (`src/data/mockData.ts`); no backend is wired up yet.