My story

I'm an AI Engineer at Tata Consultancy Services in Thane, India, where I partner directly with the CXO's office to architect agentic AI automations — work that's contributed to 25% Total Contract Value growth and a 40% reduction in executive workflow bottlenecks using Copilot Studio.

Day to day, that means designing multi-agent workflows with the Microsoft Foundry SDK and MCP tool-calling, building Graph-RAG and hybrid retrieval architectures with LangChain, and running a secure LLM proxy on Azure that serves 100+ enterprise users with sub-second response times. A Graph-RAG research effort I led was recognised among the top 15 internal research papers across a 10,000+ employee engineering division.

Before AI engineering, I trained as a mechanical engineer at the University of Mumbai, with hands-on stints in industrial robotics (KUKA robotic arms), CAD/PLM (NX11, Teamcenter), and precision manufacturing. That background shows up in how I think about software systems — as things with tolerances, failure modes, and maintenance cycles, not just clever demos.

Outside my day job, I write about the tools I use daily — LangGraph, MCP, prompt engineering, RAG design — as a guest writer and on this site's blog and notes, and I keep a research interest in astronomy, having taken part in the International Astronomical Search Collaboration's asteroid-hunting program.