I build compound AI and multi-agent systems that model customer behavior across search, selection, and econometrics at Amazon Stores Science. My research spans LLM-agent reasoning, agent reward modeling, RL from AI feedback, knowledge-graph reasoning, and graph–language model integration, with work published at ICLR, NeurIPS, KDD, WWW, WSDM, IJCAI and one US patent. Previously, I completed my Ph.D. at Virginia Tech with Dr. Chandan Reddy, and my Masters & Bachelors at IIIT, Hyderabad.