Model interpretability and observability
The Project aims at creating interpretable text representations and LLM evaluation and observability tools
PhD in Computer ScienceNorth Carolina State University
I am a Data Scientist at LexisNexis, where I work on Nexis+AI’s Retrieval-Augmented Generation (RAG) system. My current work focuses on context engineering and evaluation for conversational agentic RAG pipelines that process large-scale, real-time news data to deliver actionable insights.
Prior to joining LexisNexis, I completed my Ph.D. in Computer Science at NC State University. My academic advisor was Distinguished Professor Munindar P. Singh, and my dissertation combined language models and multiagent social simulations to analyze online news discourse and its influence on social media users. I completed my Master’s thesis under the advisement of Dr. Collin F. Lynch . My current research focuses on three key areas: Model Interpretability and Observability, Multi-Agent Social Simulations; and AI Alignment.
At NC State University, I have been an active member of the AI in Society group, helping to organize and host several of its early seminars. Additionally, I have contributed to the development of teaching materials for graduate-level courses and served as the President of the Indian Graduate Student Association (IGSA). My contributions have been honored by NC State University with awards such as Outstanding Teaching Assistant (2021) and Outstanding Student Leader (2023) .
In addition to my current role at LexisNexis, I have gained diverse experience across multiple industry positions. I previously served as a Member of the Technical Staff at Oracle for two years, and completed several Machine Learning Engineer internships at leading organizations, including Lenovo, Seagate, and Coupang, where I contributed to a range of applied AI projects. Additionally, I worked as a Research Fellow at the Laboratory of Analytic Sciences during the 2024 Summer Conference on Applied Data Science (SCADS).
The Project aims at creating interpretable text representations and LLM evaluation and observability tools
The Project aims to use multiagent simulations to understand LLM behavior across varying social contexts
The Project aims to understand the influence of news via social media analysis
I was born in Varanasi (formerly Benares), and spent my formative years there. I moved to Bangalore, India, for my undergraduate studies and later to the United States for my master's and PhD.
Beyond research and work, I am passionate about music and am a self-taught flutist. I enjoy reading nonfiction, particularly biographies and books on human behavior, psychology, and topics that explain how things work. I am a nature enthusiast and love hiking and exploring the outdoors (my latest expedition is to hike in all US National Parks, and I have covered 12 so far).
Here's my resume.