Skip to content

Blog

Notes from building and running backends, cloud platforms and AI products in production. Each post starts from something real: a migration that went sideways, a design that held up under load, a cost that surprised everyone.

Written for engineers who want the specifics, with a clear takeaway for the managers and founders who have to make the call.

PostgreSQL Isolation Levels Explained with Python

A tested Python tutorial: see what each PostgreSQL isolation level lets through, overspend a budget three ways, then stop it with SERIALIZABLE retries, a row lock or one UPDATE.

PostgreSQL24 min read

Distributed Locks in Redis with Go

A tested, step-by-step Go tutorial: reproduce two requests spending the same balance, fix it with a Redis lock, then make the fix hold up when things go wrong.

Distributed Systems29 min read

AWS Security Checklist for Small Teams

An AWS security checklist for small teams, built on the AWS Startup Security Baseline: 19 controls in three tiers, each with a command, a verify step and its cost.

AWS27 min read

GitHub Actions OIDC Trust Policy for AWS

GitHub now issues two OIDC subject formats. Here is a trust policy that accepts both, a way to prove it against every repository, and a safe order for retiring keys.

AWS14 min read

SCP vs RCP vs Declarative Policies in AWS

Deny a verb or assert a state? A practical guide to choosing between SCPs, RCPs and declarative policies, with examples and a safe rollout checklist.

AWS13 min read

Move an AWS Account to Another Organization

The order of operations for moving an AWS account between organizations: invite from the destination, enroll in Control Tower, then close the sign-on, guardrail and old-role gaps.

AWS12 min read