pgmpy: Open-Source Infrastructure for Causal AI

pgmpy is an open-source Python package for causal AI, the branch of machine learning concerned with cause-and-effect relationships rather than prediction alone. This ability is central to explainable AI and to decision-making, and is critical in fields such as healthcare, drug discovery, public policy, econometrics, risk management, and recommendation systems. pgmpy implements the probabilistic graphical modelling frameworks that underpin modern causal methods and supports tasks ranging from causal discovery and parameter estimation to probabilistic inference, causal effect estimation, and simulation. Developed under open governance and a permissive licence, and designed as a general-purpose library, it already serves as reusable infrastructure for a range of other projects. This project will broaden pgmpy's algorithmic coverage with recent methods from research and add dedicated modules for root cause analysis and policy optimization, reinforcing its role as a transparent, reproducible, and community-governed building block for causal AI.