Most enterprises are still shoehorning AI into structures built for a different era, the horseless carriage problem. In this episode, Shawn is joined by Cobus Greyling, AI Evangelist at Kore.ai, and Carl Katz, VP of Global Technology Partners, for the show's first-ever dual-guest conversation. Together, they cover why multi-agent orchestration is outpacing prompt engineering as the critical enterprise skill, how to govern a workforce of AI agents before sprawl sets in, what the shadow AI economy is costing organizations in attribution and compliance, and how technology leaders can make the business case for AI experimentation without a clean ROI to show for it. If you're trying to move urgently but thoughtfully on AI, this one is for you.

What You'll Learn:

• Why multi-agent orchestration is becoming the skill that matters more than prompt engineering and what that shift means for your team
• What "bounded autonomy" is and why setting the right autonomy level per use case is the difference between a useful AI agent and a liability
• How the shadow AI economy works against your attribution, compliance, and AI governance efforts
• Why 80–85% of enterprise AI pilots fail in-house and what the ones that succeed do differently
• The horseless carriage trap: why plugging AI into existing processes without reimagining them is a recipe for wasted investment
• How to frame the AI experimentation business case for your board and finance team when clean ROI isn't available yet
• What an agent control plane is and why you'll need one before your AI agent count gets out of hand
• How to think about the build vs. buy vs. partner decision as agentic AI moves from pilot to production