Field Notes

Notes from the field.

Short essays on building, hard-won lessons, and the occasional opinion I can't keep to myself. No cadence, no fluff, just what I've actually learned.

TEVV is where AI programs live or die

Everyone wants the model. Almost nobody budgets for proving it works. In high-assurance settings, that is exactly backwards.

Retrieval-augmented generation is the most under-appreciated tool in the applied AI toolkit, if you build it with discipline.

I left the option of a bigger firm to build a one-person AI practice. Here is the logic, for anyone weighing the same move.