On this episode of the Scaling Tech Podcast, Arin sits down with Senior Web Developer and long-time AgilityFeat and WebRTC.ventures team member Andrea Phillips for a grounded, practical look at what responsible agentic development actually looks like in the wild.
Andrea walks through her approach to building production software with AI agents, explaining why a task without a clear scope is just a “wish” waiting to go wrong, and demonstrates how she orchestrates a team of specialized AI agents for PM, development, and code review tasks. She also gets real about the risks: from vibe coding shortcuts that skip the engineering fundamentals, to cautionary tales like Meta’s high-profile Instagram security fiasco, which she uses to illustrate exactly what happens when humans get taken too far out of the loop.
If your team is adopting agentic development and wants a clear-eyed view of how to do it responsibly, without burning tokens, breaking production, or losing your architectural judgment, this Scaling Tech Podcast is essential listening.