Skip to content

DEPLOYMENT.md

Latest commit

 

History

History
107 lines (81 loc) · 5.41 KB

DEPLOYMENT.md

File metadata and controls

107 lines (81 loc) · 5.41 KB
Raw

Aarini Production Deployment Guide

This repository contains a full-stack health application with an Expo React Native frontend (compiled for Web) and a Flask Python backend.

1. Prerequisites & Environment Variables

Make sure you have accounts and projects set up on:

Backend Environment Variables (Render)

Variable Name Purpose Example / How to Obtain
FLASK_ENV Run mode production
GEMINI_API_KEY Empathetic health assistant key Obtain from Google AI Studio
FIREBASE_SERVICE_ACCOUNT_JSON Firebase authentication & db credentials Minified single-line JSON of your Service Account Key

Frontend Environment Variables (Vercel)

Variable Name Purpose Example / How to Obtain
EXPO_PUBLIC_API_URL Endpoint of backend API E.g. https://aarini-backend.onrender.com (no trailing slash)

2. Firebase Database Setup

  1. Open the Firebase Console and create a project named Aarini (or select an existing project).
  2. Enable Firestore Database:
    • Navigate to Firestore Database in the left menu.
    • Click Create Database. Select production mode and select a region close to your user base.
  3. Enable Authentication:
    • Navigate to Authentication in the left menu.
    • Click Get Started, then select Email/Password as a provider, enable it, and save.
  4. Generate Service Account Private Key:
    • Go to Project Settings (gear icon in the sidebar) > Service Accounts.
    • Click Generate New Private Key and download the JSON file.
    • Convert Key to Single-Line JSON: Open the downloaded JSON file, copy its contents, and minify it into a single line (remove line breaks). This will be used as the value for the FIREBASE_SERVICE_ACCOUNT_JSON environment variable in Render.

3. Backend Deployment (Render)

The repository includes a render.yaml Blueprint file at the root. You can deploy the backend using either the blueprint or manual configuration.

Option A: Automatic Deployment via Render Blueprint

  1. Go to the Render Dashboard.
  2. Click New + and select Blueprint.
  3. Connect your GitHub repository.
  4. Render will automatically read the render.yaml file.
  5. In the settings, fill in the values for the environment variables:
    • GEMINI_API_KEY
    • FIREBASE_SERVICE_ACCOUNT_JSON
  6. Click Apply to deploy the services.

Option B: Manual Web Service Setup

  1. Go to Render Dashboard > New + > Web Service.
  2. Connect your repository.
  3. Configure the following service settings:
    • Name: aarini-backend
    • Environment: Python
    • Root Directory: backend
    • Build Command: pip install -r requirements.txt
    • Start Command: gunicorn app:app
  4. Expand the Advanced section and add the required environment variables:
    • FLASK_ENV=production
    • GEMINI_API_KEY
    • FIREBASE_SERVICE_ACCOUNT_JSON
  5. Click Create Web Service.
  6. Once deployed, note down the service URL (e.g. https://aarini-backend.onrender.com).

4. Frontend Deployment (Vercel)

Vercel will build the Expo project into a static web application and serve it with client-side SPA routing.

  1. Go to the Vercel Dashboard and click Add New > Project.
  2. Import your GitHub repository.
  3. Configure the project settings:
    • Framework Preset: Select Other (or leave default).
    • Root Directory: Select frontend.
    • Build Command: npm run build
    • Output Directory: dist
    • Install Command: npm install --legacy-peer-deps (Crucial for handling peer dependency trees with React 19).
  4. Add the following Environment Variable:
    • Key: EXPO_PUBLIC_API_URL
    • Value: Your deployed Render backend URL (e.g. https://aarini-backend.onrender.com)
  5. Click Deploy.
  6. Once the build is complete, Vercel will provide your live frontend URL (e.g. https://aarini-frontend.vercel.app).

5. Verification Checklist

  • Health Endpoint: Visit the backend URL https://your-backend.onrender.com/ in your browser. Verify it returns {"status":"healthy", ...} and firebase_connected: true.
  • Onboarding Screen: Open the Vercel frontend URL, verify the Splash screen loads, and click sign up.
  • Database Connection: Register a test user (e.g., test@aarini.com) and verify that a user document appears under the users collection in the Firebase console.
  • AI Assistant: Navigate to the chat tab, send an inquiry about menstrual cramps, and verify it returns a response generated by Gemini AI.