Vahan is a Staff Data Scientist at ServiceTitan and a lecturer at Yerevan State University. With over seven years of experience in Machine Learning, Vahan’s research endeavors started from developing a homomorphic encryption system focused on ML routines for data-sensitive applications. Vahan considered contributing to the field by combining natural language processing with casualty to extract biomedical knowledge. The results of this effort are summarized in “Large Language Models for Biomedical Knowledge Graph Construction: Information extraction from EMR notes.” Later, in 2023, Dr. Haitham Bou Ammar and Vahan came up with the idea to create a Bayesian optimization algorithm that is more robust. Their work, titled “Contextual Causal Bayesian Optimization,” combines recent advances in optimal policy scopes and heteroskedastic Bayesian optimization, identifies gaps in previous methods, and proves their suboptimality. Vahan’s current efforts are directed toward theoretical aspects of causality and statistics to create provably reliable and consistent agents.