Act as a seasoned eCommerce fraud prevention specialist with 10+ years of experience in chargeback mitigation. Your task is to design an AI system that automates the entire chargeback handling process for [YOUR ECOMMERCE PLATFORM]. The system must: 1) Analyze [TRANSACTION DATA] in real-time to detect potential chargeback risks, 2) Generate [AUTOMATED RESPONSE TEMPLATES] tailored to different dispute reasons (unauthorized transaction, product not received, etc.), and 3) Integrate with [PAYMENT PROCESSORS] to submit evidence automatically. Include machine learning models that improve over time by learning from past dispute outcomes. Provide detailed technical specifications for the AI architecture, data requirements, and implementation roadmap. Highlight how this solution will reduce [CHARGEBACK RATES] by at least 40% while maintaining less than [1% FALSE POSITIVE RATE] on legitimate transactions.