Building enterprise-grade React applications, AI-powered interfaces, and digital experiences that are fast, scalable, and crafted with precision.
Three years in, I've spent most of my time on software that people are paid to use — aviation revenue systems, learning platforms for schools across India, fintech tooling, and now an AI agent platform that places and answers real calls.
That work rewards a particular discipline: architecture that survives the second year, component systems a team can move quickly inside, and state that stays honest when the data is streaming in live. I care about the render path as much as the layout.
The craft side matters too. Motion, timing and restraint are how an interface signals quality before anyone reads a word of it — which is roughly the argument this page is making.
From full-stack intern to owning frontend architecture on an AI platform — each step traded scope for depth.
Lead the frontend of an AI agent platform running inbound/outbound calls, SMS and email — knowledge bases, workflow automation and live conversation in one console.
A learning management system used by schools across India — complex academic logic, and a performance pass felt by an entire classroom at once.
A Solana-based platform taken end to end — wallet-aware flows, transaction states and on-chain data, plus a reusable component library for the team.
Where the fundamentals were built: full-stack apps in React, Node, Express and MongoDB, and the debugging instinct that comes from owning something end to end.
Commercial products built with teams at Verveo, Digiblock and Navneet Toptech. Screens are placeholders — the architecture, the ownership and the outcomes are not.
A platform that runs inbound and outbound voice, SMS and email on behalf of businesses. Agents are configured, grounded in a knowledge base, and set loose on real conversations — with a live console watching every one.
Built from zero as lead frontend for US-based clients: lease management, rent collection, revenue tracking and the reporting layer airports run their commercial operations on.
An end-to-end Solana product — Next.js interface, Node services behind it. Wallet-aware flows, transaction state and on-chain data presented so that nothing about it feels experimental.
Used by schools across India. Complex academic business logic, dynamic content management, and a performance pass that mattered because a slow load means a stalled classroom.
Open to frontend and AI-interface roles, and to teams who care how the thing feels. Bengaluru or remote.