Build a causal ML pipeline end-to-end

Imagine you have completed a causal-inference study in an analytics or experimentation setting and now need to deploy it as a production system that decides which individuals receive a treatment. Address the following five requirements. Build a causal DAG from treatment to outcome, add the principal confounders, and defend the conditional independence assumptions implied by that graph. Specify the estimands you care about: the average treatment effect and the…