DevOps & Cloud

This hub covers deploying and running software in the cloud: serverless functions, infrastructure as code, CI/CD pipelines, DNS and certificates, databases that scale to zero, and keeping the bill small. Most examples use AWS, because that is where Shipworthy runs, but the ideas carry over to other providers.

It is for developers who own their own deploys. That means solo builders, small teams without a platform team, and application developers who have been handed "the infra" and want to understand it rather than copy a template and hope. You do not need to be a cloud specialist; you do need to be willing to read a config file.

The approach is real configs from real apps. Code and configuration come from running systems, with the versions stated, because cloud tooling breaks between releases more often than its documentation admits. Costs come from actual bills or from official pricing pages with the date checked, never from "it's basically free". Posts call out the gotchas that cost time, and they say when a managed service is worth paying for instead of building it yourself. Security basics, such as least-privilege roles and no long-lived keys in CI, are treated as part of the job, not an appendix.

Where to start: how Shipworthy runs Next.js 16 on AWS with SST walks through the complete serverless setup behind this site, with CloudFront, Lambda, S3 and a single DynamoDB table defined in one config file, and GitHub Actions deploying through an OIDC role. Then read idempotent writes in DynamoDB with a guard item, which shows how to make a write safe to retry. That matters as soon as Lambda, a queue or a client starts retrying your requests, which on serverless is a question of when, not if.