Notes on Working Well With AI Coding Tools
For engineering managers and founders deciding whether to hire, augment a team, or just teach the developers they already have to use AI tools properly. Not tied to one language — tied to the judgement question underneath.
Nothing Published Here Yet
This is a new site and the blog is genuinely empty right now. We would rather tell you that than fill the page with three placeholder headlines that go nowhere, or posts written to hold a keyword rather than to say something.
The first pieces will come from the same material we use to run the marketplace: how to tell a developer who uses AI tools well from one who just uses them fast, what actually shows up in a careful review of AI-generated code, and why “train the team” is sometimes the better answer than “hire someone new”. When they exist, they will appear here — not before.
Until then, the vetting and hiring pages already cover most of this ground in more detail than a blog post would.
Four Subjects We Keep Coming Back To
Not stack tutorials. The questions that actually decide whether an AI-assisted engagement goes well.
Speed Versus Judgement
Where AI-tool speed genuinely helps, where it quietly hides a mistake, and how to tell which is happening from the outside.
Reviewing AI Output
What a careful review of AI-generated code looks for that an ordinary code review skips, and why confident-looking code is the risky kind.
Hire, Augment, or Train
How to tell which of the three actually fixes the problem you have, and why the answer is often the one nobody in the room suggested first.
Vetting Without Guesswork
What a dual-axis evaluation actually tests for, why the two halves are kept separate, and what a weak result on each one tends to look like.
Skip the Reading. Send Us the Brief.
Tell us the project and stack, and get matched developer profiles within days — interviewed by you, with a replacement at no cost inside the first two weeks if the fit is wrong.