Blog
Platform engineering, hybrid infrastructure, and operational discipline, from the platform outward.
Terraform Is Not a Platform (And That's Okay)
Terraform is an execution engine, not a platform. Understanding the difference and what to build above the tool layer is the starting point for platform engineering.
Read →The Future of Infrastructure Is Control Surfaces
Infrastructure tooling exposes configuration primitives, not stable operational actions. The teams that operate complex environments reliably are the ones who built a control surface above the tools.
Read →What Makes an Infrastructure Platform Coherent
Tool lists do not explain how a platform works. A coherent infrastructure platform separates intent, execution, governance, and implementation so operators have a stable interface above the tool layer.
Read →Why AI Increases the Need for Platform Engineering
AI generates working infrastructure code fast. It does not generate operational infrastructure. As generation accelerates, the platform governance layer matters more, not less.
Read →Evidence-Driven Operations
Absence of signal is not evidence of health. What operational evidence actually looks like: structured run records, inline probes, and verification that outlasts the engineer who ran the job.
Read →Disaster Recovery as a Governance System
The technical side of DR is largely solved. The harder problem is governance: who decides to use it, based on what information, and what evidence does the decision produce.
Read →Why Hybrid Infrastructure Still Matters
For universities, regulated enterprises, and public sector organisations, fully-cloud is often the wrong architectural answer. Hybrid infrastructure is not a transitional state; for many environments it is the correct long-term model.
Read →Source-of-Truth Automation with NetBox
Configuration drift is inevitable when intent is scattered across multiple systems. How NetBox as a single authoritative source changes what you can confidently know and reliably automate about your infrastructure.
Read →Teaching Infrastructure Through Outcomes Instead of Tools
Tool competence and system understanding are not the same thing. Training infrastructure engineers through operational outcomes, including scenarios, probes, and run records, produces better results than teaching tools first.
Read →Why Infrastructure Platforms Must Be Reproducible
Reproducibility means the same inputs produce the same environment, every time. Why having infrastructure-as-code is not enough, and why this property matters most when you need it most.
Read →VyOS on Hetzner as a Static WAN Hub for Dynamic On-Prem Connectivity
How inverting the IPsec initiation direction eliminates the dynamic on-prem IP problem, and why running VyOS on both sides of the link simplifies everything else.
Read →