A tech executive leveraged ChatGPT as a real-time decision engine to challenge his father's emergency amputation prognosis. By extracting clinical data, decoding medical jargon, and identifying specialized clinics under time constraints, the LLM enabled a high-risk revascularization that saved the foot—at the cost of one toe. This case study reveals a 'Good Practice Loop' for augmenting human judgment with AI under pressure.