Architect AI and Agentforce adoption across the enterprise

• AI use case portfolio prioritized by ROI feasibility and risk tier • Shared Agentforce action library and prompt standards via CoE • Human-in-the-loop for high-risk decisions financial medical • Data readiness assessment before AI—not garbage in garbage out • Trust Layer and compliance review for each production agent • Change management: agents augment staff not replace without plan • Measure quality metrics not only deployment count