Fred Mare, a principal engineer at Xero, joins Ganesh Datta to explore why production readiness should be framed around the customer experience rather than internal operational concerns. Fred shares Xero's approach to confidence scores—a signal that aggregates quality metrics, change size, and test coverage to predict the impact of any deployment on end users.

Fred and Ganesh discuss criticality tiering based on customer jobs to be done, why automation is essential to avoid overloading engineering teams, the inseparable relationship between security and production readiness, and why the best advice for starting a production readiness program is to start small and treat it as a journey.

This week, Cortex co-founder and CTO Ganesh Datta discusses the evolution of platform engineering with Kaspar von Grünberg, CEO and co-founder of Humanitec. Kaspar draws a sharp contrast between the "artisanal" method of software development and the industrialized approach required for modern enterprises. He argues that relying on individual heroes to "YOLO" their way through deployments is unsustainable, and that true scale demands standardized, reliable systems.

Ganesh and Kaspar unpack why standardization doesn't kill creativity, the critical difference between "rigid paths" and "Golden Paths," and why treating your platform as a product is non-negotiable. They also discuss the emerging role of AI, arguing that AI agents are effectively a new class of "junior engineer" that makes robust platforms more essential than ever.

Nathen Harvey, who leads the DORA research program at Google Cloud, joins Ganesh Datta to explore how AI is reshaping software delivery—and why the fundamentals matter more than ever. Nathen shares insights from DORA's 2025 report on AI adoption, the surprising finding that AI acts as an amplifier of both good and bad practices, and the critical shift teams must make from chasing elite performance to driving elite improvement.

Nathen and Ganesh discuss value stream mapping as a planning tool for 2026, why SLOs are becoming more important in an AI-first world, the trust paradox in AI adoption, how to use AI as a catalyst for cultural change, and why measuring lines of code generated by AI misses the point entirely. They also explore how foundational practices like working in small batches, strong version control, and user-centric focus become even more valuable when AI is writing more of your code.