Systems Thinking is the Agentic Unlock with Google's LaSean Smith

Episode notes

In this episode of the Shift AI Podcast, LaSean Smith, Product and Growth Lead at Google Cloud, joins host Boaz Ashkenazy for a wide-ranging conversation on how systems thinking and agentic AI are reshaping the way individuals, small businesses, and enterprises operate. LaSean shares a career journey that spans Microsoft HoloLens, Amazon, a successful startup exit, and now Google - plus a portfolio of small businesses that have served as his real-world AI lab. From a salad shop in Renton to a pre-construction development business in Seattle, he's applied workflow design and agent automation to solve practical problems long before it was fashionable. The conversation digs deep into how to actually build effective AI agents - not by prompting a chatbot, but by thinking in workflows first, identifying where reasoning actually needs to happen, and writing skills that make agents fast, reliable, and token-efficient. LaSean explains the "parcel grader" agent he built for his construction business, why he starts every agent build in a chat interface before moving to CLI, and how the McDonald's SOP model is the right mental framework for getting great output from AI.

• AI, Security, and the Courage to Reinvent with Smartsheet CTO Cynthia Tee66 / 100
• Building the AI-First City with San Jose Chief Innovation Officer Stephen Caines
• Unprompted Intelligence in High-Stakes AI with Hebbia CTO Aabhas Sharma
• The Agentic Platform Hiding in Plain Sight with Slack Senior Director of DevRel Kurtis Kemple
• AI Agent Deployment and Real ROI with MongoDB Field CTO Pete Johnson