Platform design, data engineering, AI in production, cloud architecture, and the everyday craft that holds it all together. Written by the people on the keyboard, not the marketing team.
How we built a fully serverless FinOps platform at cloudandthings.io — the architecture, the event volumes, and the design decisions that keep thousands of nightly events flowing without a server to babysit.
Combine Amazon EKS with the regional reach and compliance of Azure — running Kubernetes worker nodes across both clouds under one control plane, no second cluster required.
If you're developing with Python, Terraform, AWS, Azure, or really anything — managing your tools and environments can quickly become a mess.
Data pipelines often aren't built with the latest and greatest tech, and the brief is usually to stay within limitations while still saving costs as much as possible.
Where document processing pipelines slow to a crawl — and how an AI-powered approach breaks the bottleneck and keeps data moving.
Inside the design and delivery of some of the largest cloud landing zones in South Africa — the account structure, guardrails, and FinOps that scale to an enterprise estate.
What it takes to build and maintain PCI DSS-compliant cloud infrastructure in AWS — the controls, the evidence, and the automation that keeps you audit-ready.
Engineering for South Africa's two-pot retirement system — the software development challenges of delivering a regulated financial change on a hard deadline.
A straightforward guide to adding SSL to a single-instance Elastic Beanstalk NGINX application — simplifying secure web hosting without over-engineering it.
Hard-won lessons on achieving success in challenging software development projects — how we keep delivery on track when the brief, scope, or constraints fight back.
A simple Lambda that watches CloudTrail for manual console actions (ClickOps) and alerts the team on Slack — so your infrastructure-as-code stays the source of truth.
Turning hairy Terraform plans into something you can actually review — rendering tf plans as readable, filterable HTML so the important changes don't get lost in the noise.
Deep, multi-part writing on the topics we go back to. Updated as the work evolves — not abandoned after part two.
From a blank AWS Organization to sixteen workload accounts in production. Module shapes, guardrails, FinOps, the bits the docs skip.
Migrating a real warehouse off Hive. Performance, ops, catalog choices, and how we kept the analytics team out of meetings.
Evals, guardrails, fallback chains, observability. The full stack of habits that keeps a production LLM out of the news.
Field notes from our engineers, straight to your inbox. Unsubscribe anytime. We'll never share your email or send anything else.