Profile

I'm Shaurya Tiwari, an AI Software Engineer with a Master's in Computer Science from Florida State University.

My direction is robot learning through simulation: building manipulation environments, training policies, and studying how learned skills can retain, transfer, and reuse across tasks, objects, and domains instead of being relearned from scratch. That work runs through InstruX, a decision layer I'm building over Isaac Sim, Isaac Lab, and OpenUSD for the robot-policy lifecycle: diagnosing failures, generating simulation scenarios, and triggering retraining when deployed systems drift.

Day to day, at U.S. News & World Report I build production AI infrastructure (agentic RAG systems, backend platforms, vector retrieval, and streaming APIs) with the guardrails and observability that keep AI products reliable across several business verticals. That systems discipline is what I'm now applying to simulation and robot-policy evaluation.

Before this I worked across data engineering, research, DevOps, and web platforms, which is where I picked up the backend, ML, and infrastructure judgment I lean on today.

Toolbox

The stack I reach for, by area:

Beyond Work

Off the clock I'm occasionally corny on my YouTube livestream, figuring life out with LLMs.