Platform Engineering Monthly — October 2025

Oct 28, 2025

Welcome to the twenty-third edition of Platform Engineering Monthly! Suggestions or ideas for the next edition? Let me know!

The AWS Outage That Broke the Internet — News that few could miss: the huge us-east-1 AWS outage that took down dozens of major services, including Alexa, Fortnite, and Snapchat. There’s no shortage of analysis — from smart-beds gone rogue, to the risks of single-region dependence, to speculation on Amazon’s strategic brain drain.
GitHub Will Prioritise Migrating to Azure Over Feature Development — GitHub (and LinkedIn) have attempted and failed to do this in the past, so now the corporate overlords are requiring GitHub to dogfood Azure over serving customers. Perhaps just acquiring everyone will be Microsoft’s long term strategy to increase Azure market-share?
Why Your Platform Engineering Career Is Really a Sales Job — A sharp observation that good platform engineers aren’t just builders — they’re sellers. Success hinges on convincing developers to adopt your platform, not forcing them to.
Why Up to 70% of Platform Engineering Teams Fail to Deliver Impact — Another article on why “Platform as a Product” coupled with compelling metrics is the way to do platforms. Most teams falter due to weak adoption, unclear value, and misaligned goals.
Ulysses’ Odyssey: Lessons for Platform Engineering — Fun little read, where Greek mythology meets platform engineering in a metaphor-laden piece on resilience and iteration.

Migrating from AWS to Hetzner — With the recent AWS outages, I figured it was worth sharing yet another “I moved all my stuff to Hetzner and saved $$$”.
Applying RBAC to Databases on Kubernetes: Practical Real-World Examples — Walk-through on applying Kubernetes RBAC principles to database workloads — ideal if you’re managing Postgres or MySQL inside clusters.
Spec-Driven Development Using Markdown as a Programming Language — I like the idea of spec driven development to deliver small projects to reduce the amount of non-deterministic behaviour in AI, so this was a nice guide from GitHub.
Building a 1-Million-Node Kubernetes Cluster — Building a 1 million node k8s clusters, because why not?

Flightcontrol — A Heroku-like PaaS that deploys directly into your AWS account, providing simplicity without giving up control.
k7 — A lightweight orchestrator for running isolated VM sandboxes — think mini-VMs for reproducible environments.
Replik8s — A modern open-source Kubernetes auditing tool focused on security and policy enforcement.
Sparky — A flexible and minimalist CI server.

KubeCon + CloudNativeCon North America 2025 — Atlanta, Georgia · 11–14 November 2025
Keynotes, CNCF project updates, and plenty of platform-engineering sessions — including OpenTofu Day and BackstageCon NA.
Platform Engineering Executive Roundtable (KubeCon NA) — Atlanta, Georgia · 11 November 2025
Hosted by Platform Engineering Org — invite-only roundtable for leaders discussing the evolving role of internal platforms.
AWS re:Invent 2025 — Las Vegas · 1–5 December 2025
The cloud giant’s annual conference, with sessions on observability, AI integration, and platform automation.
QCon San Francisco 2025 — 17–21 November 2025
Tracks include platform engineering, resilient architecture, and large-scale systems design.

Have platform engineering tips to share? Reply to this email or connect with me on LinkedIn.

Platform Engineering Monthly