Curious about how AI behaves once it leaves the notebook.

I’m Chahak Goswami, an Artificial Intelligence student at Penn State. I’m most interested in the engineering layer between a capable model and a dependable system: orchestration, tools, retrieval, state, evaluation, safety gates, observability, and recovery.

My projects tend to start with a practical failure mode — schema drift, weak citations, repetitive ticket triage, broken CI — and ask where an AI agent can help without hiding risk.

I also care about cloud security and governance because autonomous systems inherit the permissions, infrastructure, and operational assumptions around them.