How would you design a system to prevent unintended medical advice from being generated in Assort Health’s patient-facing AI?
