Open to workBackend EngineerFull-Stack Engineer

I build backend systems that hold up under real load.

Ex-Antino Labs, where I owned core backend systems for Stackd — a smart vending platform serving 375K+ registered users — designing REST APIs, optimizing PostgreSQL at scale, and re-architecting order workflows end to end.

Systems worked on: 375K+ users · 22K+ MAU · 10K+ daily orders

My impact

0%faster order APIs
0%faster reporting queries
0%faster order completion
Explore work→
Let's talk

Open to backend, full-stack, and startup engineering roles.

<Selected Work/>

Personal projects built and shipped in competitive tech fests

Top 7 · Tech FestMar 2024

Hichat

Real-time chat & video calling platform

A scalable communication platform with real-time messaging, group chats, file sharing, and audio/video calling.

Node.jsReact.jsExpress.jsMongoDBSocket.ioAWS
Live →GitHub
Top 10 · College CompetitionNov 2023

DeepLink

Professional networking platform

A career-networking platform with rich profiles, job search, skill endorsements, and real-time one-to-one messaging.

Node.jsReact.jsExpress.jsMongoDBSocket.ioAWS
Live →GitHub

<Tech Stack/>

Backend

NestJSNode.jsExpress.jsREST APIsSocket.ioTypeORM

Databases

PostgreSQLRedisMongoDB

Languages

TypeScriptJavaScriptSQLC/C++

Frontend

React.jsReact Native

Tooling

GitGitHubPostmanVS Code

<Engineering Principles/>

How I actually approach building systems, not just what I've used

Measure before optimizing

Core

I profile first, guess later. The 87% latency cut on Stackd's order API came from tracing the actual bottleneck — the inventory reservation step — not from broad, speculative changes.

Keep systems observable

Core

You can't fix what you can't see. Reporting queries that silently took 6+ minutes only got fixed because the execution flow was actually inspected end to end, not assumed to be fine.

Build for scale, not for show

Indexing, pagination, and execution-plan tuning across 5+ high-traffic endpoints mattered more than any single clever query — scale comes from consistent discipline, not one-off tricks.

Simple beats clever

Async processing over synchronous reservation logic wasn't a clever trick — it was the obvious fix once the bottleneck was clear. The simplest fix that addresses the real cause wins over a complex one that impresses.

<Work Experience/>

Associate Software Developer

Most recent role

Jun 2025 - Jul 2026

Antino Labs • Gurugram

Owned core backend systems for Stackd, a smart vending platform serving 375K+ registered users and 22K+ MAU.

87%faster order APIs
98%faster reporting queries
58%faster order completion

The 87% latency cut came from parallelizing the inventory reservation step — moving it off the request path into async processing instead of resolving it synchronously per order.

  • Redesigned reporting queries and the database execution flow, cutting API response time from 6+ minutes to 5–6 seconds
  • Re-architected the end-to-end order workflow, cutting average completion time from ~2min to ~50s
  • Designed REST APIs handling 10K+ daily orders across inventory, wallets, refunds, and vending machine operations
  • Tuned indexing, pagination, and execution plans across 5+ high-traffic endpoints to reduce database load
  • Shipped features across web and mobile (React.js / React Native), partnering with backend teams end-to-end
NestJSPostgreSQLRedisTypeORMNode.jsReact.jsReact Native