Define platform architecture requirements with Engineering and Architecture - including data models, ingestion patterns, APIs, extensibility, security, and scalability - and decide where AI capabilities and MCP services are the right way to deliver platform value versus traditional APIs and integrations. Working knowledge of AI and generative AI (LLMs, retrieval, agents), including practical trade-offs in quality, cost, latency, and trust, with an AI-first mindset for designing capabilities usable by agents, people, and conventional integrations.