Accelerating Value with Agentic AI

Mark Relph, Director of Data and AI Partners at AWS, explores the rapid evolution of AI from generative tools to Agentic AI—autonomous software entities capable of reasoning and completing tasks independently—predicting that by 2028, 30% of enterprise applications will be powered by these systems. He highlights the rise of task-specific AI models, such as those supported by Amazon Bedrock and Amazon Q, and emphasizes their transformative impact on workplace productivity, citing examples like the NFL and Moody’s. Relph underscores the critical importance of secure, responsible AI implementation, supported by robust data foundations and clear guardrails. He introduces Amazon Nova’s suite of foundation models—Nova Micro, Lite, Pro, Premier, NovaReal, and NovaCanvas—which offer cost-effective, high-performance solutions optimized for multi-agent and task-specific scenarios, boasting up to 75% greater efficiency. Relph concludes with optimism about the agentic future, affirming AWS's commitment to powering this future through trusted infrastructure, specialized tools, and strategic partnerships with companies like Domo.