Payments, data & systems that cannot go down

Before building a platform for shop owners, I worked on systems where downtime means real money lost — a nationwide AI-powered chatbot handling 100,000+ active sessions, automating financial services, branch discovery, and live agent support with co-browsing and video.

That experience taught me how to build for reliability — WebSocket infrastructure, microservices, Disaster Recovery drills, and production debugging under pressure. The same discipline goes into every payment flow, every stock update, and every invoice generated on this platform.

Business logic built from the ground up

Spent a year building a management system for state and district-level sports authorities — attending events on-site, working directly with organizers, and translating messy real-world operations into clean backend workflows.

That on-ground experience — watching how people actually work, not just how they say they work — directly shaped how this platform handles walk-in billing, wholesaler records, and multi-role access. Real operations rarely match the spec. The platform is built knowing that.

The platform you're looking at, right now

Everything on this page — the billing flows, the GST engine, the B2C storefront, the wholesaler credit system — is a production-grade multi-tenant SaaS platform built from scratch on Laravel 12 and Next.js, with every schema, service, and business rule designed specifically for how Indian retail actually works.

GST-compliant invoicing with CGST/SGST/IGST splits. Event-driven inventory via Redis queues with atomic stock deduction. Razorpay subscriptions with race condition handling. AI-powered product suggestions via Google Gemini. Not prototypes — production code.

Where it started

BE in Electronics & Communication Engineering, University of Burdwan — CGPA 8.68. The engineering mindset — systems thinking, data structures, networks, operating systems — is what separates a platform built to scale from one that breaks when things get busy.