Cloud cost engineering, plainly.
Specific, numbers-heavy writing on what an AWS bill actually looks like, when Hetzner saves you 70 percent, where multi-cloud arbitrage pays off, and the pricing model traps nobody warns you about. No 101 content.
Pricing model traps: per-second billing, prorated months, minimum charges
Per-second billing isn't always what it sounds like. Prorated months on Hetzner. AWS minimums. The billing-unit details that distort apples-to-apples comparison.
The total cost of self-hosting a Kubernetes cluster on each provider
Managed EKS/GKE/AKS vs self-hosted k3s on each cloud. Control plane fees, node group economics, ingress controller cost, and the real all-in number.
Cloud TCO formulas that aren't lies (with a worked spreadsheet)
TCO formulas that include the line items vendor calculators conveniently omit — committed compute, real egress, ops time, migration cost, opportunity cost.
Cost of running a single web app on each provider
Same Rails/Django/Node web app: 2 web instances + Postgres + Redis + 500 GB egress. Real 2026 list prices across 7 providers.
Cost of running a small Postgres on each provider in 2026
Same Postgres workload — 2 vCPU, 8 GB RAM, 200 GB SSD, daily backups, high availability — priced across AWS RDS, GCP Cloud SQL, Hetzner, DO, Vultr, Linode.
When to leave AWS for Hetzner, Vultr, or DigitalOcean
Specific signals that you will save money by moving off AWS: egress profile, managed-service dependence, regulated workload status, team operational capacity. Decision checklist.
Network bandwidth pricing decoded: per-GB, per-Mbps, per-flow
Per-GB egress, included monthly quotas, port speed caps, inter-AZ charges. Six provider models compared on the same workload pattern.
Reserved capacity vs committed-use: the contract details that matter
AWS RI scope rules, GCP CUD exchange policy, Azure RI cancellation policy. The fine print that decides whether your commitment is worth signing.
Burst credit instances (T-family, e2, B-series): when they save money, when they hurt
AWS t3, GCP e2, Azure B-series — the burst-credit pricing model is brilliant for some workloads and a hidden tax for others. Worked examples both ways.
Multi-cloud arbitrage in practice (with a worked example)
Real worked example for a mid-sized SaaS: stateless tier on Hetzner, managed Postgres on AWS RDS, CDN on Cloudflare. 60% cheaper than single-cloud AWS.