Vishal Verma
I am a Research Engineer and ML Leader specializing in causal inference, distributed ML systems, and LLM infrastructure. With over 10 years of experience spanning industry and academia, I drive applied AI innovation at the intersection of rigorous research and scalable engineering.
My research focuses on causal representation learning—uncovering latent causal structures from observational data to enable more robust, interpretable, and trustworthy machine learning systems. I am particularly interested in how causal abstractions can improve generalization, fairness, and out-of-distribution robustness in modern ML pipelines.
Currently, I am advised by Prof. Kun Zhang (CMU/MBZUAI) and Prof. Zhijing Jin (Max Planck Institute), working on foundations of causal discovery and representation learning. Previously, I led data science initiatives at Dream11 as Lead Data Scientist, where I built production ML systems serving millions of users.
Get In TouchReleased fully automated causal effect estimation library powered by Large Language Models.
"Causal AI Scientist: Facilitating Causal Data Science with Large Language Models" accepted at Neurips Workshop.
"Causal AI Scientist: Facilitating Causal Data Science with Large Language Models" accepted at Conference on Language Modeling.
Contributed distributed computing capabilities to scale OrthoLearner algorithms in Microsoft's EconML library using Ray framework.
Papers on FENCE fraud detection system and distributed causal algorithms accepted at ACM AIML conference.
Upreti, Akriti; Verma, Vishal; Kothari, Kartavya; Thukral, Utkarsh
Read PaperVerma, Vishal; Reddy, Vinod; Ravi Jaiprakash
Read PaperPittsburgh, Pennsylvania