I'm an AI/ML engineer who likes going a little too far down technical rabbit holes.
I enjoy building things that are not just working, but make me stop and ask: "Okay... but what happens if this breaks?"
That question has led me into multi-agent systems, RAG, LLM evaluation, guardrails, model routing, observability, and a frankly unhealthy amount of documentation.
- 🤖 AI systems that can actually survive the real world
- 🔍 Understanding what happens under the hood
- 🧪 Breaking systems so I can make them harder to break
- 🛠️ Building tools because "someone should make this" is usually enough reason
- 🧩 Turning weird ideas into surprisingly serious projects
- 🥇 Microsoft Agent League — Global Student Award
- 🥇 Gappy AI Hackathon
- 🏅 Top 5 — Techathon 2.0
- 🏅 AIR 13 — National Innovation Challenge for Drone Application & Research
Learning, building, breaking, rebuilding.
If something makes me think "wait, can we actually do that?" I'm probably already trying it.
- Portfolio → nithisha.vercel.app
- Email → nithishaleni1806@gmail.com
- LinkedIn → linkedin.com/in/nithisha-v