DEPLOYMENT.md

🚀 CodeScan AI - Deployment Guide

This guide covers deploying CodeScan AI to production using Vercel (frontend) and Render (backend).

---

📋 Prerequisites

Before deploying, ensure you have:

• GitHub repository with CodeScan AI code
• Vercel account (https://vercel.com)
• Render account (https://render.com)
• API keys for:
  • Groq API (https://console.groq.com)
  • Google Gemini (https://makersuite.google.com/app/apikey)
  • Hugging Face (optional, https://huggingface.co/settings/tokens)

---

🎨 Frontend Deployment (Vercel)

Step 1: Connect GitHub Repository

1. Visit https://vercel.com/new
2. Click "Import Git Repository"
3. Select your CodeScan AI GitHub repository
4. Vercel will auto-detect it as a Vite project

Step 2: Configure Build Settings

Build Command:

npm run build

Output Directory:

dist

Root Directory:

frontend

Step 3: Set Environment Variables

In Vercel Project Settings → Environment Variables, add:

VITE_API_URL=https://api.codescan-ai.com
VITE_SOCKET_URL=https://api.codescan-ai.com
VITE_APP_NAME=CodeScan AI

For Development:

VITE_API_URL=http://localhost:5000
VITE_SOCKET_URL=http://localhost:5000

Step 4: Deploy

Vercel will auto-deploy on every push to main branch. You can also manually trigger deployments from the Vercel dashboard.

Your frontend will be available at:

https://codescan-ai.vercel.app

---

⚙️ Backend Deployment (Render)

Step 1: Create Flask API Service

1. Visit https://dashboard.render.com
2. Click "New +" → "Web Service"
3. Connect your GitHub repository
4. Configure the service:

Name: codescan-api

Environment: Python 3.10

Build Command:

pip install -r requirements.txt

Start Command:

gunicorn --worker-class eventlet -w 1 run:app

Root Directory:

backend

Step 2: Add Environment Variables

In Render Service Settings → Environment, add:

# Flask Config
FLASK_ENV=production
SECRET_KEY=your-super-secret-key-here

# Database
DATABASE_URL=sqlite:////var/data/codescan.db

# AI Provider Keys
GROQ_API_KEY=your_groq_api_key
GEMINI_API_KEY=your_gemini_api_key
HUGGING_FACE_API_KEY=your_huggingface_token

# Redis (will be added separately)
REDIS_URL=redis://default:password@redis-instance.onrender.com:6379
CELERY_BROKER_URL=redis://default:password@redis-instance.onrender.com:6379/0
CELERY_RESULT_BACKEND=redis://default:password@redis-instance.onrender.com:6379/1

# CORS Settings
FRONTEND_URL=https://codescan-ai.vercel.app
CORS_ORIGINS=https://codescan-ai.vercel.app,http://localhost:5173

# JWT Config
JWT_SECRET_KEY=your-jwt-secret-key
JWT_ACCESS_TOKEN_EXPIRES=900
JWT_REFRESH_TOKEN_EXPIRES=604800

Step 3: Add Persistent Volume

1. In Render dashboard, go to Service Settings
2. Add a Disk volume:
   • Name: data
   • Mount Path: /var/data
   • Size: 1 GB (or more for production)

This ensures your SQLite database persists across deployments.

Step 4: Add Redis Add-on

1. In Render dashboard, click "Add-ons"
2. Create a Redis instance:
   • Name: codescan-redis
   • Plan: Free (or paid for production)
3. Copy the Redis URL to REDIS_URL environment variable

Step 5: Deploy

Render will auto-deploy on every push. Your API will be available at:

https://codescan-api.onrender.com

---

🔄 Celery Worker Service (Background Tasks)

Create Celery Worker Service

1. In Render, create a new Background Worker
2. Configure:

Name: codescan-worker

Build Command:

pip install -r requirements.txt

Start Command:

celery -A celery_worker.celery worker --loglevel=info --concurrency=2

Environment Variables: (Same as Flask API service)

Connect to Redis

The worker will automatically connect to the Redis instance using the REDIS_URL environment variable.

---

🗄️ Database Setup

SQLite (Development/Small Scale)

Already included. No additional setup needed.

PostgreSQL (Enterprise/Production Scale)

For larger deployments, migrate to PostgreSQL:

1. In Render, add PostgreSQL Add-on
2. Update DATABASE_URL environment variable
3. Run migrations:

flask db upgrade

---

🔒 SSL/HTTPS

Both Vercel and Render provide automatic SSL certificates. HTTPS is enabled by default.

---

📨 Email Notifications (Optional)

To enable email notifications for scan results:

1. Set up SendGrid integration in Render
2. Add to environment variables:

SENDGRID_API_KEY=your_sendgrid_api_key
SENDGRID_FROM_EMAIL=noreply@codescan-ai.com

---

🔐 Security Checklist

Before going to production:

• Set strong SECRET_KEY and JWT_SECRET_KEY
• Enable HTTPS (automatic on Vercel/Render)
• Configure CORS to allow only your frontend domain
• Set up API rate limiting
• Rotate API keys regularly
• Set up monitoring and alerts
• Enable database backups
• Use environment variables for all secrets
• Set up error logging (Sentry/LogRocket)

---

📊 Monitoring & Logging

Render Logs

View logs in Render dashboard:

Service → Logs

Error Tracking (Optional)

Integrate Sentry for error tracking:

import sentry_sdk
from sentry_sdk.integrations.flask import FlaskIntegration

sentry_sdk.init(
    dsn="your_sentry_dsn",
    integrations=[FlaskIntegration()],
    traces_sample_rate=0.1
)

---

🔄 CI/CD Pipeline

GitHub Actions Workflow

Create .github/workflows/deploy.yml:

name: Deploy to Production

on:
  push:
    branches: [ main ]

jobs:
  deploy-frontend:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Deploy to Vercel
        run: |
          npm install -g vercel
          vercel deploy --prod --token ${{ secrets.VERCEL_TOKEN }}

  deploy-backend:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Deploy to Render
        run: |
          curl https://api.render.com/deploy/srv-${{ secrets.RENDER_API_ID }}

---

🐛 Troubleshooting

API Not Starting

Error: ModuleNotFoundError: No module named 'flask'

Solution:

pip install -r requirements.txt

Database Connection Error

Error: sqlite3.OperationalError: unable to open database file

Solution: Ensure /var/data volume is mounted and writable

Redis Connection Timeout

Error: redis.exceptions.ConnectionError

Solution: Verify Redis URL in environment variables and network connectivity

Socket.IO Connection Issues

Error: WebSocket connection fails

Solution:

1. Ensure FRONTEND_URL is correctly set
2. Check CORS configuration
3. Verify WebSocket is enabled in Render (default: enabled)

---

📈 Scaling

Horizontal Scaling (Multiple Instances)

For production load, use Render's auto-scaling:

In Render Service Settings:

• Enable "Auto-Deploy"
• Set instance count to 2-3
• Configure load balancer (included)

Upgrade Database

Migrate SQLite → PostgreSQL:

# Export data
sqlite3 codescan.db .dump > backup.sql

# Create PostgreSQL instance on Render
# Update DATABASE_URL
# Run migrations
flask db upgrade

---

💰 Cost Estimation

Component	Free Tier	Pro Tier	Notes
Vercel Frontend	100 GB bandwidth/month	$20/month	Includes SSL, CDN
Render API	$7/month	$12+/month	Auto-scales, includes SSL
Render Redis	Free	$5+/month	Managed Redis
Render PostgreSQL	N/A	$15+/month	For scaling
Celery Worker	$7/month	$12+/month	Background jobs

Total Estimated Cost: $7-40/month depending on scale

---

📞 Support

• Vercel Support: https://vercel.com/support
• Render Support: https://render.com/docs
• CodeScan AI Issues: https://github.com/pauldev-hub/CodeScan-AI/issues

---

🔄 Updates & Maintenance

Zero-Downtime Deployments

Both Vercel and Render support zero-downtime deployments:

1. New version is deployed to a canary instance
2. Health checks pass
3. Traffic gradually shifts to new version
4. Old version is terminated

Rollback

If deployment fails:

Vercel: Go to Deployments, select previous version, click "Make Production"

Render: Go to Deploys, select previous deployment, click "Deploy"

---

✅ Post-Deployment Checklist

• Frontend loads at custom domain
• API responds to health checks
• Authentication works
• Scans complete successfully
• Results export (PDF, JSON, CSV)
• Real-time chat works
• Error logging is active
• Backups are scheduled
• Monitoring/alerts are set up
• Performance is within SLA