Halo.mp4

Halo is a healthcare operations and monitoring platform that combines outbound patient calling, doctor workflows, and a nurse-facing 3D hospital digital twin in one integrated stack.

How It Works

Upload: Import a hospital floor plan (PNG/JPG) and reconstruct room layout locally in the browser.

Visualize: Explore the floor in a persistent Three.js scene with priority-coded care rooms and occupancy state.

Assign: Sync live patient and room data from the backend so only occupied care rooms become interactive.

Inspect: Open a patient room to view a 3D body model with condition markers mapped to clinical severity and body area.

Operate: Switch to Doctor mode to manage patients, author call questions, and launch outbound check-in calls.

Monitor: Track call session progress and history while backend webhooks orchestrate turn-by-turn phone workflows.

Iterate: Update patient records and room assignments; the 3D scene refreshes from the latest backend feed.

Quick Start

Prerequisites

• Node.js 18+
• Python 3.10+
• MongoDB (local or Atlas)
• API keys for Twilio Voice (outbound calls)
• Optional: Gemini API key (live room generation and condition guidance)

Frontend Setup

cd frontend
npm install
npm run dev

Open http://localhost:3000.

Set the backend URL before Doctor mode will connect correctly:

DOCTOR_API_BASE_URL=http://127.0.0.1:8000

Backend Setup

cd backend
python -m venv venv
source venv/bin/activate   # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env       # add MongoDB URI, Twilio, and service credentials
uvicorn main:app --host 0.0.0.0 --port 8000

Architecture

Frontend

• Next.js and React: App Router UI with Nurse and Doctor modes
• Three.js: Interactive 3D hospital floor plan and patient body visualization via React Three Fiber and Drei
• Next API routes: Server-side proxy to backend (/api/doctor/*) and live room-data merge endpoints

Backend

• FastAPI: Patient CRUD, session history, call initiation, and Twilio webhook orchestration
• MongoDB: Persistent storage for residents, sessions, and call artifacts
• Twilio Voice: Outbound calls with HTTP webhook-driven conversation state machines
• Speech and reasoning pipeline: Transcription, acoustic analysis, and structured clinical summaries for check-in workflows

Inspiration

Care teams often split their work across phone workflows, spreadsheets, and disconnected monitoring tools. Halo brings those flows into one system: doctors can run structured outreach from a dashboard, while nurses can see who needs attention spatially on a hospital floor instead of parsing static lists.

The goal is to make remote patient check-ins and room-level clinical awareness faster, clearer, and easier to act on in real time.

License

MIT