Skip to content
for recruiters
AS00 · Arrival

Open to backend & full-stack roles

Abhinav Shah

Node.js / MERN Stack Developer

I design and build backend systems that stay calm under load — clean APIs, deliberate data models, and full-stack MERN products shipped end to end.

View selected workGet in touch
Find meGitHub ↗LinkedIn ↗
scroll to begin ↓

Who I Am

A backend mind, full-stack reach.

Abhinav Shah — Node.js / MERN Stack DeveloperOpen to backend & full-stack roles
Based in
Ahmedabad, Gujarat · India
Timezone
IST (GMT+5:30)
Languages
Hindi, English

I'm a backend-leaning full-stack developer focused on scalable Node.js services and the MERN stack. I care about the parts users never see: data modeling, authentication, rate limiting, and the quiet reliability that lets a product grow without rewrites.

Currently studying Computer Science and building production-grade services — the kind of work that values clarity over cleverness. I like code that tells a story: every function with a purpose, every module worth reusing.

10k+
API requests / day handled
8.71
CGPA · B.Tech CSE
2
flagship MERN builds
∞
curiosity for systems

// skill map

Languages & DBs

  • TypeScript
  • JavaScript (ES6+)
  • Python
  • SQL
  • PostgreSQL
  • MongoDB
  • Redis

Frameworks & Libs

  • Node.js
  • Express.js
  • Next.js
  • React.js
  • Electron.js
  • tRPC
  • Prisma
  • Drizzle ORM

Tools & Architecture

  • RESTful APIs
  • Microservices
  • Turborepo
  • WebSockets
  • Docker
  • Git/GitHub
  • CI/CD
  • Cloudflare

B.Tech in Computer Science Engineering

Indus University

2022 – 2026CGPA: 8.71

How I Think

Principles before frameworks.

Tools change. The way you reason about systems is what compounds. Here's the lens I bring to every build.

  1. 01

    Model the data first

    Most bugs are data-shape bugs in disguise. I design schemas and contracts before endpoints, so the API is a thin, honest layer over a sound model.

  2. 02

    Make failure boring

    Rate limits, validation, and idempotency aren't extras. Predictable failure modes keep a system debuggable at 2am and at 10,000 requests a day.

  3. 03

    Optimize the bottleneck, not the vibe

    Measure before tuning. A well-placed index or aggregation beats a clever rewrite. Performance is a budget, not a guess.

  4. 04

    Ship small, ship reversible

    Incremental, reversible changes over big-bang rewrites. Every deploy should be easy to roll back and easy to reason about.

// tech radar

Adopt

  • Node.js
  • Express
  • MongoDB
  • React
  • TypeScript
  • REST

Trial

  • PostgreSQL
  • Docker
  • BullMQ
  • Redis

Assess

  • GraphQL
  • tRPC
  • Kubernetes
  • gRPC

Selected Work

Two builds, examined.

Fewer projects, told properly — the problem, the architecture, the trade-offs, and what shipped. Try the live endpoint that serves this section:

01

NEXUS

A next-gen smart traffic control dashboard. Optimizes complex 4-lane intersections in real-time with AI and computer vision.

Read case study
2026
Live Source

Problem

Modern cities are choking on static traffic lights. Urban planners need dynamic systems that can "see" congestion and adapt on the fly, but combining live simulation with raw computer vision is notoriously difficult to scale.

Approach

I engineered a high-performance Express & React platform capable of digesting live WebSocket feeds. It intelligently swaps between a synthetic traffic engine and a Python-based YOLO/OpenCV microservice that tracks vehicles frame-by-frame.

architecture

  • Vite/React dashboard with 2s WebSocket pulse
  • Python CV microservice running YOLO & OpenCV for raw video analysis
  • DataSourceManager pattern for seamless simulation vs. live-video toggling

results

  • Achieved sub-2s latency for live intersection rendering
  • Integrated an AI Copilot for natural-language traffic querying
Latency
<2s
AI
Copilot
  • React
  • Express
  • Vite
  • YOLO
  • OpenCV
  • PostgreSQL
  • WebSockets
02

ShipFlow AI

The ultimate AI product manager. ShipFlow orchestrates the entire software lifecycle from a raw idea to deployment-ready code.

Read case study
2026
Live Source

Problem

The gap between product ideation and engineering execution is massive. PRDs get ignored, QA reviews take days, and miscommunication causes endless delays in shipping features.

Approach

I built a comprehensive multi-tenant SaaS that acts as an intelligent bridge. It uses AI to challenge and refine product requirements, automatically generates engineering tasks, and runs a relentless autonomous QA review against GitHub Pull Requests.

architecture

  • Next.js App Router Monorepo powered by tRPC
  • Inngest durable workflows for resilient AI code reviews
  • Deep GitHub OAuth & Webhook integration for PR tracking

results

  • Fully automated the QA review loop directly inside GitHub
  • Reduced PR friction by instantly catching requirement deviations
Architecture
tRPC Monorepo
Workflows
Durable
  • Next.js
  • tRPC
  • Prisma
  • PostgreSQL
  • Inngest
  • Tailwind CSS
03

Mart

A lightning-fast, production-grade e-commerce engine engineered with a modern Turborepo monorepo architecture.

Read case study
2026
Live Source

Problem

Legacy e-commerce platforms suffer from fragmented codebases, slow page loads, and a lack of real-time interactions, resulting in poor developer experience and lost sales.

Approach

I architected a highly scalable fullstack solution from the ground up. By utilizing Turborepo, I isolated the React frontend, Express API, and shared validation schemas, creating a completely type-safe and real-time capable platform.

architecture

  • Turborepo workspaces for strict API/Web boundary enforcement
  • Shared Zod schemas & Drizzle ORM for end-to-end type safety
  • Socket.io for instant live updates and notifications

results

  • Delivered a remarkably fluid, app-like shopping experience
  • Dramatically improved DX with shared monorepo tooling
Scale
Turborepo
Real-time
Socket.io
  • React
  • Express
  • Turborepo
  • Zustand
  • PostgreSQL
  • Drizzle
  • Socket.io
04

FormCraft

A gorgeous, full-stack form builder SaaS. Think Typeform, but open, customizable, and instantly scalable.

Read case study
2026
// runs on my machineSource

Problem

Off-the-shelf form builders are expensive and rigid. Creators need a tool that offers highly interactive, one-question-at-a-time form filling with deep analytics and monetization built-in.

Approach

I crafted a drag-and-drop visual builder using dnd-kit, backed by a robust Next.js and Express architecture. It dynamically evaluates conditional logic on the client while gracefully handling heavy submission loads via Redis rate-limiting.

architecture

  • Next.js frontend with complex dnd-kit drag-and-drop state
  • Express REST/tRPC backend with Upstash Redis rate limiting
  • Razorpay integration for subscription-tier billing

results

  • Built a beautifully fluid one-question-at-a-time UI
  • Enabled deep response analytics and CSV exports
UI
dnd-kit
Billing
Razorpay
  • Next.js
  • Express
  • Drizzle ORM
  • tRPC
  • React Query
  • Upstash Redis

System Design

A request, traced end to end.

The most honest way to show how I think about backends is to follow one request through the system. This scene is rendered with Remotion — a real animation timeline, not a video.

fig.03 — checkout request lifecycle · Client → API → Auth → Queue → DB
  1. 01Client

    React app issues POST /api/checkout

  2. 02API

    Express validates input and applies rate limiting

  3. 03Auth

    JWT verified, RBAC role resolved

  4. 04Queue

    Side-effects (payment webhook) enqueued to BullMQ

  5. 05Database

    Order persisted; 201 returned in ~142ms

Decoupling side-effects through a queue keeps the request fast and the system resilient — webhooks and emails never block checkout.

Playground

Don't take my word for it — poke the backend.

Real endpoints, running live. Hammer a rate limiter, ask the portfolio a question, or open the terminal (⌘J) and explore by command.

›press ⌘J for the full terminal — try ls, ask, query stack=React, github, sudo

Experiments

Small probes, real lessons.

Things I build to understand a system better — not products, but proofs.

Rate-limit playground

backend

Token-bucket vs. sliding-window, visualized.

Aggregation studio

data

Composing MongoDB pipelines step by step.

Auth flow sandbox

security

JWT refresh rotation with replay protection.

Queue inspector

infra

Watching BullMQ jobs move through states.

Open Source

Learning in public.

I learn in public — small, focused repositories and contributions over grand unfinished monorepos.

18

Public repos

640

Contributions / yr

6

Languages

lessmorecontribution cadence · illustrative

Setup

Behind the scenes.

The environment I think in. Optimized for fast feedback loops and as few context switches as possible.

~/.config/dev.toml
  • editorVS Code · Vim keybindings
  • terminalWarp · zsh + starship
  • runtimeNode.js LTS · nvm
  • osWindows + WSL2 / Ubuntu
  • keyboard65% mechanical · tactile
  • fuelCoffee, measured in PRs

// daily toolchain

  • Docker
  • Postman
  • MongoDB Compass
  • GitHub
  • Jira
  • Figma

$git commit -m "ship small, ship reversible"
$npm test && npm run deploy

// deployment workflow

Let's Build

Let's build something reliable.

Open to backend and full-stack roles, and to interesting problems. The fastest ways to reach me:

Emailabhinavshah133@gmail.com
Phone+91-7878625157
GitHubgithub.com/Shahji-01
LinkedInlinkedin.com/in/abhinavshah133