Hej — I'm Jens.
Independent platform and infrastructure engineer, London. Twenty-five years on GCP and AWS, building the layer underneath — landing zones, infrastructure as code, observability, security and cost — and the code on top of it. I have done this for seed-stage startups with a handful of engineers, and for enterprises where a twenty-person function reported to me. The job is the same shape at both ends: make the safe thing the easy thing, then hand it over.
Platform foundation for a subscription business
Sole platform engineer on a consumer SaaS: organisation landing zone in Pulumi, factory components giving every service observability, cost guards, IAM and secrets by default.
Operating model for an infrastructure team
Interrupt-driven support replaced with a chat-based intake on a Kanban board, and documentation standards raised so engineers resolved the common cases themselves.
Security, cost and release engineering in insurance
Zero standing access: Workload Identity Federation for CI, entitlements for humans, no service-account keys. CI/CD runtime cut by 60%.
What I actually do in the first month on a Google Cloud estate that grew without one — why I don't start with Terraform, why the billing console comes before the architecture diagram, and what has to be true before I leave.
Terraform owns the landing-zone literature and Pulumi does not, so the GCP-at-org-scale questions go unanswered. Factory components, state layout, ephemeral per-PR stacks, and the mistake I made first.
Google Cloud, Firebase, Cloudflare, Neon, Astro, React, Swift and Kotlin. The stack I build solo apps on, why I chose each piece, and why it suits my situation — not yours by default.
A clip about apps that print money — and the stop-vaping example in it — made me see the gap between an income and a business, and rethink where my hours go.
This site uses privacy-friendly Google Analytics to see what's read — no ads, no cross-site tracking. It stays off unless you accept.