I am a full-stack developer from Kolkata, India with experience building and shipping products across frontend, backend and cloud environments.

I like projects where the product is visible and the engineering matters: real-time tracking, interactive interfaces, async workflows, ML-assisted features, secure execution environments and production deployment.

Right now, my flagship project is SkyWatch Live — an airspace intelligence platform that tracks live aircraft, visualizes satellites and surfaces unusual flight behaviour using a three-model anomaly detection pipeline.

SkyWatch Live is a real-time airspace surveillance and anomaly detection platform built with public aviation and orbital data.

At the captured project benchmark, it tracked 11,862 active aircraft including 10,894 airborne aircraft on 30-second polling cycles, combining state vectors from 5+ public feeds with cross-source deduplication.

• Ingested live ADS-B data from multiple public sources, including OpenSky, ADS-B One and FLARM, and displayed unique aircraft on an interactive Leaflet canvas map.
• Implemented a three-model ML ensemble — Isolation Forest, Local Outlier Factor and MLP Autoencoder — over a 30-dimensional engineered feature space for asynchronous trajectory anomaly detection.
• Indexed 85,390 airport facilities across 249 countries for origin and destination resolution.
• Added SGP4 satellite propagation using CelesTrak TLE data, alongside live FAA TFR and AWC SIGMET airspace overlays.
• Designed the event-driven workflow using Django Channels, Celery and Redis so ingestion and ML scoring run asynchronously while telemetry streams over WebSockets.
• Exposed Prometheus metrics with custom gauges, counters and histograms for a Grafana observability dashboard.

View the repository →  ·  Read the architecture →  ·  Read the build story →

A full-stack coding practice platform built around safe code execution, structured problem tracks and AI-assisted learning.

• Built a custom judge engine that executes Python and Java submissions inside isolated Docker containers rather than running arbitrary user code on the host.
• Designed validation for exact-match and floating-point outputs across 853 problems spanning 35+ topic tracks.
• Added a Monaco-based editor for an IDE-like browser experience.
• Integrated Razorpay payment webhooks with idempotent event handling to avoid duplicate subscription grants.
• Added Gemini-powered hints that help users progress without revealing full solutions.

React Node.js PostgreSQL Docker AWS Gemini API

Dec 2024 - Sep 2025 · Kolkata, India

• Served as the sole developer on Inca, a live ride-booking platform: built the Node.js backend, React Native Android apps for customers and drivers, and a Next.js admin dashboard; shipped the product to the Play Store where it reached 500+ downloads.
• Implemented real-time trip tracking and driver matching with Socket.io and MongoDB geospatial queries, keeping customer, driver and admin views synchronized without a polling layer.
• Built product catalogue, cart and order-management workflows for an e-commerce platform using Next.js and MySQL, giving the business a single admin interface for inventory and orders.
• Diagnosed N+1 query patterns and introduced composite indexes across MySQL and MongoDB to improve response times on frequently used endpoints without changing the schema.
• Set up GitHub Actions CI/CD pipelines, environment-based configuration management and graceful shutdown handling to prevent in-flight request drops during deployments.

Mar 2024 - May 2024 · Remote, United States

• Migrated core frontend components from Angular to React using TypeScript strict mode, eliminating a recurring class of runtime type errors in the existing codebase.
• Built asynchronous request handling with retry logic and streaming support for LLM API integrations, keeping the UI responsive during high-latency model responses.
• Brought UI components up to WCAG-AA accessibility standards, improving screen-reader usability and resolving cross-browser responsiveness gaps.

B.Tech, Computer Science Engineering
Techno India University, Kolkata · 2020 - 2024 · CGPA: 8.31 / 10

Google Cloud Digital Leader
Google Cloud · Dec 2023