In this episode, Carlos Gonzalez de Villaumbrosia interviews Chris Geoghegan, VP of Product at Zapier. As the company’s first-ever Product Manager, Chris has spent nearly a decade scaling Zapier into a $5 billion automation giant that serves over 3.4 million businesses and 69% of the Fortune 1000.
Zapier is not just building AI tools; they are powering their entire company with them. Chris reveals that his team currently runs over 800 active AI agents internally to manage everything from calendar prep to engineering triage. He breaks down the Code Red moment that shifted their strategy and how they are defining the future of Agentic Workflows.
What you’ll learn:
- Agentic vs. Deterministic: Why standard workflows follow a set path, while agents can reason, access knowledge, and change course to solve problems.
- The Orchestration Layer: How to hire and onboard AI agents using Context Engineering and Model Context Protocols (MCPs).
- Adoption vs. Transformation: Why adoption is just doing old tasks faster, while transformation unlocks business models that were previously impossible.
- Building a Moat: How Zapier uses its vast data on user intent to stay ahead of commodity LLM features.
Key takeaways:
- Treat Agents Like Employees: You can't just deploy an agent; you must onboard it with specific context and tools to be effective.
- Lead by Building: Transformation fails if leaders don't use the tools. Zapier’s execs do show-and-tell sessions to prove they are hands-on.
- AI Governance is Key: To move up-market to the enterprise, you must solve for Observability (who sent what data) and Access Control.
Credits:
Host: Carlos Gonzalez de Villaumbrosia
Guest: Chris Geoghegan
Social Links:
- Find out more about Product School here
- Follow our Podcast on TikTok here
- Follow Product School on LinkedIn here
