Deploying an AI chat app on a cloud server requires critical upfront decisions on model size, server hardware, and