Gaurav Kumar Yadav - Developer Portfolio

Production portfolio for Gaurav Kumar Yadav, focused on Python, AI/ML, and full-stack development.

Live URLs

• Portfolio: https://ggauravky.vercel.app
• Admin Panel: https://ggauravkyadmin.vercel.app
• Backend API: https://dev-portfolio-ojfs.onrender.com

Stack

Frontend:

• React 18 + Vite
• Tailwind CSS
• Framer Motion
• React Router

Backend:

• Node.js + Express
• MongoDB + Mongoose
• Gemini API integration
• Security middleware (Helmet, rate-limit, mongo-sanitize, validators)

Observability:

• pino structured logs
• Request IDs on every request
• Optional Sentry error monitoring

Architecture

Monorepo layout:

• src/: React app (pages, components, hooks, utilities)
• backend/: Express API and Mongo models
• api/: Vercel serverless handlers
• shared/: shared chatbot core + privacy utilities
• public/: static assets, robots, sitemap
• scripts/: build-time utilities (sitemap, data sync, image variants)

Chatbot logic is centralized in shared modules and reused by backend and serverless entrypoints.

Features

• Multi-page portfolio (home/about/skills/projects/blog/contact/links)
• AI chatbot with fallback logic and chat analytics
• Contact + newsletter APIs
• Project and blog detail pages with SEO metadata
• Responsive image pipeline with AVIF/WebP variants + srcset
• Build-time sitemap generation for static + blog + project routes

Local Development

Prerequisites:

• Node.js 18+
• npm 9+
• MongoDB Atlas connection string

Install:

npm install
cd backend
npm install
cd ..

Run frontend:

npm run dev

Run backend:

cd backend
npm run dev

Environment Variables

Root .env:

VITE_API_URL=http://localhost:5000

Backend .env:

MONGODB_URI=your_mongodb_connection_string
PORT=5000
PORT_RETRIES=10
NODE_ENV=development
FRONTEND_URL=http://localhost:5173

# Chatbot
GEMINI_API_KEY=your_gemini_api_key
GEMINI_MODEL=gemini-2.0-flash

# Chat analytics privacy
CHATLOG_RETENTION_DAYS=60
CHAT_HASH_SALT=replace_with_strong_random_secret

# Admin/API security
ADMIN_KEY=replace_with_secure_admin_key
RATE_LIMIT_WINDOW_MS=900000
RATE_LIMIT_MAX_REQUESTS=5
CHAT_RATE_LIMIT_MAX=30

# Observability (optional)
LOG_LEVEL=debug
SENTRY_DSN=
SENTRY_TRACES_SAMPLE_RATE=0.1

Build and Production

Build command:

npm run build

Build pipeline includes:

1. Sync chat data
2. Generate responsive image variants
3. Generate sitemap
4. Vite production build

Preview build:

npm run preview

Deployment Flow

Frontend (Vercel):

1. Import repository in Vercel
2. Set root env vars (VITE_API_URL)
3. Deploy from main branch

Backend (Render):

1. Create Web Service from backend directory
2. Build command: npm install
3. Start command: node server.js
4. Add backend env vars from section above

API Endpoints (Core)

• GET /health
• POST /api/contact
• POST /api/newsletter/subscribe
• POST /api/newsletter/unsubscribe
• POST /api/chat
• GET /api/chat/privacy-policy

SEO and Indexing

• robots.txt: public/robots.txt
• sitemap.xml: generated into public/sitemap.xml
• canonical, Open Graph, Twitter, and JSON-LD: managed via src/hooks/useSEO.jsx

Image Performance Strategy

• Source images: public/images/projects and public/images/blogs
• Generated variants: 480, 768, 1200 widths
• Formats: AVIF, WebP, and fallback originals
• Runtime delivery: LazyImage component with picture + srcset + sizes

Generate variants manually:

npm run generate:image-variants

Security Notes

• Proprietary project (not open source for reuse)
• Chat analytics store hashed network identifiers, not raw IP/user-agent
• MongoDB TTL automatically expires chat logs (30 to 90 day window)

See:

• backend/PRIVACY_RETENTION.md
• LICENSE

License

Copyright (c) 2026 Gaurav Kumar Yadav. All rights reserved.