Deployment Guide

Open CIS is deployed on Railway with separate services for the API, frontend, and databases.

One-Click Deploy

The fastest way to deploy your own instance:

This provisions the full stack: API, frontend, two PostgreSQL databases, and EHRBase.

Live Environments

Environment	URL
Staging (Web)	open-cis-web-staging.up.railway.app

Railway Configuration

Each service has a railway.toml for deployment configuration. The API Dockerfile handles migrations and optional staging data seeding:

CMD sh -c "prisma migrate deploy && \
  if [ \"$RAILWAY_ENVIRONMENT\" = \"staging\" ]; then \
    python scripts/seed_staging.py; \
  fi && \
  uvicorn src.main:app --host 0.0.0.0 --port ${PORT:-8000}"

Environment Variables

Set these in your Railway project:

Variable	Description
DATABASE_URL	Prisma connection string (auto-set by Railway PostgreSQL)
EHRBASE_URL	EHRBase REST endpoint
CORS_ORIGINS	JSON array of allowed frontend origins
RAILWAY_ENVIRONMENT	Set to staging for automatic synthetic data seeding

Staging Data

When RAILWAY_ENVIRONMENT=staging is set, the API automatically seeds synthetic data on startup:

• 15 synthetic patients with realistic demographics
• 2-5 vital signs readings per patient
• Clinically plausible values
• MRN prefix: STAGING-
• Idempotent: safe to run on every deployment

See ADR-0005: Synthetic Data for details.