Available for work
Anurag Poonia

Anurag Poonia

Full-Stack Software Engineer

Full-stack software engineer from IIT Madras building and shipping production SaaS since 2024 across Next.js, TypeScript, Node.js, and React.

about

I own the full lifecycle as sole engineer for international clients — system design, REST APIs, database schema, testing, and CI/CD deployment — with a focus on scalable architecture, security hardening, and reliable zero-to-one delivery.

Sole engineer across 4+ international clients on Upwork, $7,000+ earned with a Top Rated badge in my first 6 months. Currently a final-year B.Sc. Data Science & Programming student at IIT Madras (CGPA 8.5/10).

based in
Hisar, Haryana, India
hours
Works EST for US clients
education
IIT Madras — CGPA 8.5/10
upwork
Top Rated · $7,000+ earned
More about me
experience
  • Founding Software Engineer · Facile

    Digital Identity / Smart-NFC Platform

    Jun 2026 — Present
  • Full-Stack Software Engineer · Independent

    International Clients · Upwork, Top Rated

    Jan 2026 — Present
  • Full-Stack Software Engineer · Antspace

    US-based SaaS Startup

    Nov 2025 — Jul 2026
selected work

Built a production SaaS that tracks GitHub issues and generates AI-powered insights, with full authentication, protected routes, and scalable backend APIs.

Seeded the analysis engine with 100+ repositories and 500+ issues scraped and indexed for on-demand AI analysis, with a normalized ingestion pipeline feeding the insight generation layer.

  • Next.js
  • TypeScript
  • Supabase
  • Claude API
Read case study
StakeXWeb3

Developed a production-grade Ethereum staking dApp with UUPS (ERC1967) upgradeable smart contracts, time-based reward accrual, and an ERC20 reward token.

Implemented a Web3 frontend with wagmi + viem supporting MetaMask, WalletConnect, Phantom, and Backpack, with explicit gas handling and real-time reward tracking.

  • Solidity
  • Next.js
  • wagmi
  • viem
Read case study
WatchtowerDevTools

Engineered a production uptime-monitoring system using incident-based downtime detection, with automated HTTP health checks via secure background workers and email alerts on downtime and recovery.

Cut AI operating cost ~60% by routing tasks across models by complexity — reserving frontier models (Claude 4.8) for hard reasoning and delegating routine parsing to cheaper models (Claude 4.6, GPT-4o-mini) — plus a memory/caching layer to eliminate redundant calls without degrading output quality.

  • Node.js
  • Background workers
  • SMTP
  • Multi-model AI
Read case study
skills
  • TypeScript
  • JavaScript
  • Next.js
  • React
  • Node.js
  • Express
  • NestJS
  • GraphQL
  • PostgreSQL
  • MySQL
  • Supabase
  • Firebase
  • Google Cloud
  • Docker
  • Nginx
  • Vercel
  • Railway
  • GitHub Actions
  • Git
  • PostHog
  • Sentry
  • PM2
  • Solidity
  • Claude API
github

What I’ve been building lately — hover to bring it to life.

education & recognition
  • B.Sc. Data Science & Programming

    Indian Institute of Technology (IIT) Madras

    2023 — 2026 · CGPA 8.5 / 10
  • Diploma in Programming

    IIT Madras

    CGPA 8.7+
  • Diploma in Data Science

    IIT Madras

    CGPA 8.7+
  • Top Rated

    Upwork — $7,000+ earned in the first 6 months

    4+ international clients
available for freelance

Let’s build something solid.

Open to full-stack engineering work for international clients. The fastest way to start is right here on Upwork — send a short brief and I’ll get back to you.