From requirement to running system.

I build backend and AI systems: RAG pipelines, multi-agent and tool-calling agents, and the data infrastructure underneath them. At Design Intelligence LLP I work across FastAPI services, LangChain/LangGraph agent workflows, and the ETL and storage pipelines that feed them — QuickBase-to-SAP integration, S3/Wasabi asset processing, proxy layers for credential isolation.

What I actually care about is system design: taking a loosely defined requirement and turning it into an architecture, then a working product. Deployment is part of owning that end to end, not the point of it on its own — I use Docker, Dokploy, and Vercel to ship what I build, and I'm comfortable with CI/CD and Kubernetes, but I'm not positioning myself as a dedicated deployment engineer. I'd rather be the person who can take a requirement and carry it the whole way through.

Two side projects push that further: TechPluse, a fully automated RAG and LLM pipeline that discovers, deduplicates, and summarizes AI research daily with no manual step; and MyLedger, where I designed the data architecture, security rules, and inventory logic end to end and shipped it to a live, real-world user.

Before software, I worked as a maintenance engineer on solar plant systems — different domain, same instinct: a system either holds up under real conditions or it doesn't.

BTech, Computer Science and Engineering, MAKAUT / Elitte College of Engineering, 2023 — 2026