Richardson, Texas · The University of Texas at Dallas
PhD student in Computer Science
About me
I am Ranveer, a PhD student working with Dr Sriraam Natarajan in the Statistical Artificial Intelligence and Relational Learning Group (StARLinG Lab) at the University of Texas at Dallas.
I have called UT Dallas home since coming here for my undergraduate degree in Computer Science in Fall 2017, and I continued directly into the PhD program right after finishing my undergraduate degree.
As an undergraduate, I interned at a financial company in India, where I worked on statistical modelling tools for mutual fund investment.
Research
My research currently involves large language models, probabilistic models, relational learning, reinforcement learning, and AI for healthcare. I investigate the use of LLMs for downstream tasks across clinical domains, ranging from relational information extraction from electronic health records to evaluating the role of LLMs in causal reasoning to solving relational multi-agent problems.
Exploring whether LLMs can serve as approximate knowledge sources to support learning in clinical settings, which are often data-poor and developing methods and frameworks for evaluating and mitigating hallucination and stochastic behavior in clinical decision-support tasks.
Studying multi-agent domains where states and actions are defined by objects and their relations, creating richer but exponentially larger problem spaces.
Selected publications
An automated pipeline that extracts clinically relevant variables from audio recordings to construct a Bayesian Network and reason about Alzheimer's disease progression markers.
Read paper →BibTeX ↓Causal modelling on a tiny dataset of rural or Indigenous populations living with dementia.
Read paper →BibTeX ↓Evaluating LLMs for causal reasoning in medicine.
Read paper →BibTeX ↓Combining LLMs with data refinement on data of pediatric patients on ECMO to learn a causal Bayesian network.
Read paper →BibTeX ↓Solving relational multiagent domains using MaRePReL.
Read paper →BibTeX ↓The difference in hypotension and hypertension for patients on ECMO across different sampling rates.
Read paper →BibTeX ↓The workshop version of the manuscript published at CODS 2025.
Read paper →BibTeX ↓The workshop version of the MaRePReL framework published later.
Read paper →BibTeX ↓Work experience
The University of Texas at Dallas · Computer Science Department
Working as an RA as part of my PhD. I have been funded by the NIH, primarily focusing on clinical AI in data-scarce domains.
The University of Texas at Dallas · Computer Science Department
Worked as a Teaching Assistant for CS 4141 Digital Logic Lab, conducting lab sessions, teaching and demonstrating experiments, and grading student work. Instructed over 800 students over two years.
Fintricks Consultancy Services · Delhi, India
Developed statistical models to select better mutual fund policies using temporal modelling of high-frequency stock data. Also converted MS Excel-based workloads to SQL for better integration with Python and other scripting languages.