Welcome to the SOTA Advisory writing hub

Most AI advice on the internet is written for someone else. Either it's tuned for the model researcher deciding between two training strategies, or it's tuned for the marketing lead trying to make a case for an AI feature. Very little of it speaks to the person actually deciding whether to spend the next $200K on inference.

This hub is for that person.

What will show up here

Short essays, most under a thousand words, on the questions I keep answering in engagements:

• Where AI actually changes gross margins, and where it's just table stakes.
• How to instrument an AI feature before it ships so you can trace every dollar to a user outcome.
• Where AI belongs in a workflow and where it introduces risk you can't accept.
• What Interim CTO transitions look like when the roadmap can't wait for a full-time hire.
• The gap between a strategy that looks good in a deck and one that survives the codebase.

Why these, and not more of the same

Because most of what's out there is a slide deck. My job as fractional CTO is to close the gap between the deck and the shipping product, and the writing here reflects that. Every post starts from something I actually shipped, or something I saw fail to ship, and works backward to the decision that shaped it.

If any of that would be useful for your team, book a call.