// What we build
If it can't deploy, it isn't a deliverable.
Buying a product does not transfer the obligation. Responsibility for an AI decision is non-delegable, and somebody still has to build the governance file. We build it with the model rather than after it.
// What ships
Eight artifacts, every build.
Roughly half the states have adopted the NAIC AI Model Bulletin and the variations are material. Virginia replaced “mitigate the risk” with eliminate the risk. Connecticut requires an annual AI compliance certification attested by a named officer. Iowa formally defines bias and outcomes testing.
NAIC and state bulletins · current at September 2026.
New York DFS states it plainly: an insurer “may not rely solely on a third-party’s claim of non-discrimination or a proprietary third-party process to determine compliance with anti-discrimination laws.” That sentence is the whole reason this is a build and not a procurement question.
// What's hard about this
The QA seat is the line item nobody funds.
If you deploy an extraction system and nobody owns reviewing its output, you have moved work rather than removed it.
Insurance operating-role postings barely mention exception handling, escalation or QA, which are the exact skills an AI deployment creates demand for. So we scope the review role as part of the build: the runbook, the correction loop, and who sits in it. Or we do not scope the build.
The seat reads as a soft cost to a CFO in a cost-cutting year. We put it in the proposal explicitly rather than hiding it in a blended rate, and it has cost us work.
Bring us one workflow.
We will tell you what your compliance file is missing and what has to be built to fill it. How the read works →