Deployment Overview

This guide covers deploying your chat application to production. You have several options depending on your budget and requirements.

Deployment Options

Option Cost Effort Best For
Hugging Face Spaces (backend) Free tier Low AI/ML apps, students
Render (backend only) Free tier (sleeps after inactivity) Low Free hosting
Fly.io (backend only) ~$2-5/month Medium Always-on production
VPS (DigitalOcean, AWS) ~$6-50/month High Full control
EAS Build (mobile) Free tier (30 builds/month) Low Publishing mobile app

What You Need to Deploy

  1. Backend: Flask app → production server (Gunicorn + Docker on Hugging Face Spaces)
  2. Database: SQLite → PostgreSQL (Supabase for production)
  3. Mobile: Expo app → APK/IPA → App Store / Google Play
  4. API Keys: OpenAI (or other provider) configured on the server

For a first deployment, use this path:

Backend: Hugging Face Spaces (free, Docker-based)
  ↓
Database: Supabase PostgreSQL (free tier)
  ↓
Mobile: EAS Build → share QR code via Expo
  ↓
Final: Publish to app stores (optional)

Files in This Section

File What it covers
backend-production.md Deploy Flask to Hugging Face Spaces step-by-step
database-production.md Migrate from SQLite to PostgreSQL (Supabase)
mobile-production.md Build APK/IPA with EAS Build
cost-estimation.md How much will this cost at scale?

Prerequisites for Deployment

  • A GitHub account (to push your code)
  • A Hugging Face account (sign up at https://huggingface.co/join)
  • An Expo account (sign up at https://expo.dev)
  • Docker installed (for building the production Docker image)
  • Your project committed to a git repo
cd aiproject
git init
git add .
git commit -m "Initial commit"
git remote add origin https://github.com/YOUR_USERNAME/aiproject.git
git push -u origin main