DevOps is for Product Engineers, too.
summary
Argues that platform engineering is a sociotechnical fix for scaling organizations, not just infra, and claims SLOs plus error budget policies and observability reduce incidents; describes a real observability team that expanded into reliability management, recommends dedicated on‑call shifts so engineers don’t juggle projects and paging, warns about tech debt from an opinionated Go framework (Gizmo) at scale, and urges buying boring third‑party tools, building only when adoption and measurement justify it.
AI-written summary · about
chapters (22)
- 0:00Defining sociotechnical systems
- 0:58Joint optimization
- 2:17Platform engineering strategy
- 8:57Scaling through platforms
- 10:41DevOps principles
- 11:24The pendulum of tension
- 11:53Core principles in practice
- 14:14Measurement and sharing
- 18:07Platform architecture
- 18:24Balancing developer experience
- 20:02Design principles
- 28:43Organizational leadership
- 29:32Leadership roles
- 32:37Intervention framework
- 33:54Preventative strategies
- 39:15Proactive measures
- 39:25Process improvement
- 40:27The reactive state
- 41:00Heroism and burnout
- 41:22Consequences of heroism
- 41:38Leadership accountability
- 43:33Closing reflections
related talks
22mDevOps vs Platform Engineering: Navigating the Path Ahead
41mDevOps for Java Developers (or maybe against them?!)
47mDevOps or DevX – Lessons We Learned Shifting Left the Wrong Way
37mDevOps, 12 Factor and Platform Engineering
32mDistributed DevOps
25mDevOps: Why do we Keep Messing Up?
28mDevOps meets reality
29mDevOps at Scale is a Hard Problem