I went to Guru Nanak Dev Engineering College in Ludhiana and finished my CS degree in 2025. I started writing code that touched production halfway through — a MERN event platform during an internship: Stripe webhooks, JWT, role-based access. It worked.

It worked the way an exam answer works: complete, untested by anything weird. The weird stuff came later — when actual people showed up, and the architecture I'd written without much thought started telling me what it really thought of my assumptions.

A year at Masters' Union, mostly on the proctoring system, taught me that the gap between "passes the demo" and "survives 200 users" is where the real engineering happens.

The first time the server CPU pinned at 100% with five users on it, I learned an architecture decision can be wrong — not slow, wrong. The rewrite that followed — moving inference to client worker threads, switching to an SFU — was a lesson in the difference between optimization and reconsideration.

The messaging refactor, the analytics pipeline, the event classifier — all variants of the same lesson. Most of my year was spent there. Most of what I now believe about systems came from there.

Right now I'm pulling apart how people build at the edge of what's possible — the papers, the post-mortems, the boring parts that don't make it into talks.

The thing I find most interesting in 2026 is LLMs as infrastructure problems. Not the chat UI. The queueing, the rate limits, the failure modes when inference takes eight seconds and the user is still typing. I want to work on the systems that hold that up.

If you're building there, I'd like to hear about it.