Staff-level Frontend Engineer · Systems & Developer Experience

I build the systems and tooling that help teams move faster.

Gurugram, India · open to remote

95%

faster dataset onboarding (~5 days → 3–4 hrs)

39→3

CI jobs (infra ~4h40m → ~12 min)

10s→4s

initial bundle load

30–40%

faster deploys

01 / Selected work

01

AI coding agents & issue→PR automation

Problem —
The team spent hours on the routine steps between picking up a task and opening a PR, and could not yet rely on AI tools.
Approach —
Wrote agent rules, a set of reusable skills, and an AGENTS.md capturing the product and domain context an agent needs. Built an automated pipeline that takes a task from issue to PR — picks it up, makes the fix, runs tests, opens the PR, and reports back.
Impact —
AI tools became a dependable part of everyday development; custom skills (ship-pr, figma-to-code, babysit-pr) removed repetitive work.
  • AI coding agents
  • AGENTS.md
  • GitHub Actions
02

CI/CD overhaul & Vite monorepo

Problem —
Run-All CI ran 39 jobs and ~4h40m of infra time; deploys were slow.
Approach —
Reworked the pipelines with smart lint caching and dependency reuse; moved the app into a full-stack monorepo on Vite.
Impact —
Run-All checks 39 → 3 jobs, infra ~4h40m → ~12 min, deploys 30–40% faster.
  • GitHub Actions
  • Vite
  • monorepo
03

Tuxedo v2.0 design system

Problem —
A legacy component library blocked consistent, incremental UI modernisation.
Approach —
Designed a modular design system that runs alongside the legacy library for gradual migration; solved theming and isolation with scoped CSS.
Impact —
Teams migrated incrementally without a rewrite; consistent theming across modules.
  • React
  • TypeScript
  • scoped CSS
04

Unified dataset onboarding framework

Problem —
Onboarding a new dataset took ~5 days of cross-team effort.
Approach —
Standardised API contracts with design and backend and built reusable components the team kept using.
Impact —
Onboarding time cut ~95% — from ~5 days to 3–4 hours.
  • React
  • TypeScript
05

Screener & analytics rework

Problem —
The analytics layer needed flexible, no-code data exploration and shareable state.
Approach —
Built a configurable Screener with backend-driven filters, user-controlled columns, real-time tagging, and saved/shareable views. Added a chat interface and a visual query builder; refactored the data layer with a scalable state pattern.
Impact —
Non-technical users can slice data without code; Notion-style collaboration on persistent, shareable workspaces.
  • React
  • React Query
  • TypeScript
  • AG Grid
06

Performance modernization

Problem —
Initial bundle load averaged ~10s on data-intensive dashboards.
Approach —
Introduced Webpack chunking, route-based code splitting, strategic caching, and library trimming; built a dead-code detection plugin; handled Node, AG Grid, and Vite upgrades.
Impact —
Initial load ~10s → ~4s; improved perceived performance across dashboards.
  • Webpack
  • Vite

Personal projects

Calendar Puzzle Game

A drag-and-drop puzzle where players arrange pieces to solve a calendar-based board. Built with React.

Task Management App

A dynamic task board with multiple cards and drag-to-move tasks between them. Built with React.

02 / About

I like building the systems and tooling that help a team move faster, not just shipping features.

Over the past couple of years most of my work has helped everyone around me: getting AI tools into our everyday development, making the codebase safe for people outside core engineering to contribute to, and automating the repetitive steps between picking up a task and opening a PR.

I still care a lot about performance and clean architecture, and I spend a good part of my time mentoring engineers and improving how the team works.

03 / Experience

  1. Senior Frontend Engineer (SDE-3)

    April 2021 – Present

    Synaptic Inc, Gurugram · SDE-1 → SDE-2 → SDE-3

    • Got AI coding agents into everyday development and automated the issue→PR pipeline.
    • Led a CI/CD overhaul (39→3 jobs) and the migration to a Vite monorepo.
    • Designed Tuxedo v2.0 design system and the unified dataset onboarding framework.
    • Built the Screener + analytics rework; drove performance from ~10s to ~4s.
    • Run biweekly knowledge-sharing sessions and mentor the team.
  2. Associate Software Engineer

    June 2019 – April 2021

    Innoplexus Consulting Services, Pune

    • Cut manual React–Redux boilerplate with reusable abstractions.
    • Improved reliability with smarter caching and automatic retries.
    • Built an internal charting library used across multiple projects.

Internships

04 / Skills

Frontend

  • ReactJS
  • React Native
  • Redux
  • React Query
  • D3.js
  • Webpack
  • Vite

Languages

  • TypeScript
  • JavaScript

AI & Automation

  • AI coding agents
  • Agent rules & skills
  • AGENTS.md
  • Task-to-PR pipelines

Tools & Platforms

  • Git
  • CI/CD (GitHub Actions)
  • Testing & coverage
  • Custom ESLint
  • AG Grid
  • shadcn/ui
  • JIRA
  • Agile/Scrum

Leadership

  • Technical direction
  • Cross-team work
  • Mentoring
  • Developer experience

05 / Writing

I write about frontend architecture, React patterns, and developer experience on Medium.

Read on Medium

06 / Education & Recognition

Education

B.Tech, Computer Science & Engineering

Indian Institute of Technology (IIT) Roorkee · 2015 – 2019

Recognition

  • 2nd place, Synaptic Hackathon 2023 — social-network insights prototype.
  • Star Player of the Team, Jan 2025 — technical leadership & system-level impact.
  • Published a blog on state-management architecture in Synaptic Engineering.

07 / Contact

Let's build something.

Based in Gurugram, India and open to remote. I'm looking for staff-level roles across frontend systems, platforms, and developer experience — email is the fastest way to reach me.