4+ years shipping cross-platform products across FinTech, IoT, Social and Document Security — with a focus on offline-first architecture, real-time data, and AI-assisted delivery.
Flutter Developer
Cross-platform engineer, offline-first by instinct
I build Flutter applications that stay usable when the network does not. Most of my work has involved products where connectivity is unreliable or data arrives faster than the UI can afford to rebuild — amplifier control over Bluetooth, live stock prices, mesh messaging between nearby devices. That has pushed me toward local-first persistence, explicit sync state, and architecture that separates transport from presentation.
I have delivered 10+ production apps for global clients, including Yamaha Corporation and a platform built with PolicyBazaar. I own architecture, performance and code quality decisions, and I use AI coding agents daily as a delivery multiplier rather than a substitute for engineering judgement.
What I work with
Career timeline
Products I helped ship
Audio amplifier control application with real-time device communication over Bluetooth Low Energy and MQTT. Built to remain fully functional offline, so the app stays in sync with hardware state without a network.
Stock market application with live market data and interactive charts. The engineering challenge was sustaining high-frequency price updates without dropping frames, which meant careful rebuild scoping and render-path tuning.
Social networking ecosystem delivered across mobile and Web platforms from a shared Flutter codebase.
Membership-based worker platform, built in collaboration with PolicyBazaar.
Secure document signing platform for Android and iOS.
Professional networking application covering profiles, connections and messaging flows.
Content distribution application for mobile.
Cross-platform issue tracking application for Android and iOS.
Two personal Flutter projects backed by Supabase: PostgreSQL schema and SQL design, email OTP authentication, Row Level Security policies, and offline-first sync between Hive local storage and the cloud using pending-change queues that flush once connectivity returns.
How I build with AI coding agents
AI is part of my daily toolchain across client and personal projects. What follows is the actual process I follow, not a list of tools I have tried.
I write prompts as specifications and maintain per-project AI rule files, so generated Dart follows the existing architecture, state management and naming conventions.
Large features get split into tasks an agent can complete and I can verify, instead of one prompt that produces code nobody can review.
API models and serialization, repositories, services and state management wiring — the code that is predictable and high-volume.
Widget and unit tests are generated with AI, then I check coverage and the edge cases the model did not think to cover.
Feeding stack traces, logs and failing tests to the agent shortens the distance between a bug report and the actual cause.
Every generated change is read and refactored before it lands. Architecture, security and performance decisions stay mine.
The principle: AI output is a draft, not a deliverable.
Background
Let's talk
I am open to Flutter engineering roles. The fastest way to reach me is email, and my full résumé is available as a PDF.