Seth Earley

Everyone’s talking about ChatGPT and Large Language Models (LLMs) like they’re the silver bullet for business transformation. Spoiler alert: they’re not. Without the right data and information architecture, your AI initiatives will crash and burn—fast. In this no-BS session, Seth Earley pulls back the curtain on what executives really need to know before they throw money at the latest AI trend. Drawing from 25+ years of consulting for Fortune 1000 companies, Seth reveals the dirty little secret of AI: it’s only as good as the knowledge infrastructure you build beneath it. He’ll share real-world examples of companies that got it right (and those that didn’t), showing how AI can drive massive efficiency, boost revenue, and even save hundreds of millions—if you do it right. Want to avoid becoming the next cautionary tale? This is the session for you.

Everyone’s obsessed with AI’s shiny front-end features—chatbots, virtual assistants, generative models. But here’s the truth: without the right knowledge structures behind the scenes, your AI is just an expensive toy. In this high-energy session, Seth Earley reveals the critical piece of the AI puzzle that 90% of organizations overlook: knowledge engineering. It’s not sexy, but it’s the difference between AI that delivers results and AI that flops. Seth will share a battle-tested strategy for building a “knowledge factory” that feeds your AI the structured, high-quality data it needs to thrive. You’ll learn how to break down complex business problems into solvable data challenges, build scalable knowledge models, and integrate continuous learning processes that keep your AI sharp. If you want to stay ahead of competitors, reduce costs, and deliver better customer experiences, this is the session you can’t afford to miss.

Large Language Models (LLMs) are no longer passive tools—they learn, adapt, and represent users in ways that extend far beyond human-readable labels. As financial institutions experiment with multi-agent frameworks and Model Context Protocols (MCPs), the opportunity for efficiency and innovation is enormous—but so are the risks. Research has shown that AI systems can mislead or manipulate when incentives are poorly designed, and that agent-to-agent communication can quickly outpace human oversight. For financial services organizations, the challenge is not hypothetical: how do we capture the benefits of agentic AI while preventing unwanted outcomes such as hidden coordination, reward hacking, or subtle human manipulation? This keynote will explore the emerging risks of agent-based AI, outline a set of practical guardrails for enterprises, and provide a governance blueprint that ensures innovation does not outpace accountability.