AI and engineering
How AI changes how we build — without losing velocity or judgment.
Practical lenses on AI in the SDLC, team habits, and what still requires human ownership.
Start here
Using AI in the Software Development Lifecycle (Without Slowing Shipping)
Where AI helps in the SDLC—and where it doesn’t. How to adopt AI coding tools and automation without slowing down delivery.
The Paradox of AI-Acceleration: Why We Are Typing Faster but Shipping Slower
AI lets us generate more code—but cycle times aren't dropping. Why velocity without tests, review, and observability just adds mass and risk.
Don't ask AI to build the house. Ask it to build the scaffolding.
The real value of AI in development isn't production code—it's support code: tooling, tests, and scaffolding that gets you to the hard part faster.
More in this topic
AI-Assisted Product Engineering: Orchestrating Claude Code Across the Software Development Lifecycle
Most LLM coding tools live inside a single editor session. Real product engineering doesn't. An architecture that orchestrates Claude Code across the full software lifecycle, with Python handling the mechanical steps and the agent invoked only when judgment is required.
I'm Bullish on AI-Assisted Coding. That's Exactly Why I Take the Risks Seriously.
The loudest AI enthusiasts are also the most allergic to discussing tradeoffs. That's the real risk. Honest enthusiasm is what keeps promising technologies alive.