AI Engineering & Architecture
The new layer in software: how models run, retrieval, knowledge graphs, agents and MCP, evaluation, observability, security, cost, and leading AI delivery.
18 chapters · 85 figures · 15 code panels
Edition 1
Engineering Handbooks
For experienced engineers coming back to the field: what changed, why, and where to see it working in real code.
Three handbooks, each a single web page with a printable PDF. Every chapter is anchored to one real, public service, so each idea comes with the decision record behind it, the lines of code that implement it, and the measurements that justified it.
Modern Software Engineering
Your skills, brought up to date: the service and its repository, Linux and networks, Postgres, sign-in, the web tier, containers, testing, and running it in production.
19 chapters · 72 figures · 28 code panels
Cloud Delivery & DevOps
From a laptop to production: Git and CI/CD, the supply chain, identity without keys, infrastructure as code, AWS networking, ECS, Kubernetes, and a real deployment.
14 chapters · 43 figures · 25 code panels
One reference service
triage-assistant is an on-call assistant built as the template a team starts its AI services from: sign-in with the organization's identity provider, team-owned data under row-level security, a streamed model answer grounded in the team's alerts and runbooks, evals, drills, and a deployment to AWS through a real pipeline.
Code excerpts in the handbooks are checked line by line against the tag handbook-1 (commit 5b4f80d), and every code link points to that exact version.