At Cyient, I built a 15-node commodity symbol mapping pipeline using LangGraph, PostgreSQL, and Gemini AI for automated ingestion, classification, and mapping.
I designed human-in-the-loop review nodes with a background watchdog to cancel stale approvals and maintain audit logs. I also developed a weighted validation engine and an AI Router with rate-limit retries to help prevent batch processing failures.
On the Ward-Wise Pollution Action Dashboard group project, I helped develop a near real-time dashboard for ward-level air, water, and noise pollution. It integrated environmental datasets and citizen feedback, with an AI chatbot for ward-specific recommendations.
As Departmental Head at COSMOS, NSUT, I directed computational architecture and technical execution across departmental projects, and managed web development and deployment for society portals. I'm studying Information Technology at Netaji Subhas University of Technology (NSUT).
