Abstract: Systems/techniques that facilitate two-tiered machine learning generation of birth risk score are provided. In various embodiments, a system can access a plurality of medical feature collections associated with a pregnant patient. In various aspects, the system can generate, via execution of a plurality of first trained machine learning models, a plurality of embedded features based on the plurality of medical feature collections. In various instances, the system can compute, via execution of a second trained machine learning model, a risk score based on the plurality of embedded features, wherein the risk score indicates an amount of risk to a health of the pregnant patient or a health of a fetus of the pregnant patient that is associated with performing a caesarian-section on the pregnant patient or with waiting for the pregnant patient to give birth naturally.