What we have learned building this in production
Four things we write about, because they are the four places AI projects actually go wrong: the deployment question, the process underneath, the regulation around it, and everything that only breaks once real users arrive.
Your AI pilot will fail its security review
The pilot works. Then legal asks where the data goes, and the answer is an API in another jurisdiction with a sub-processor list nobody can enumerate. The project stops there, usually after the budget is spent.
Read the piece
The first improvement usually needs no AI
On a field sales engagement that ended with a 3–4% increase in sales, most of the gain came before a model was involved. We deleted a step instead of automating it.
Read the piece
Reading the EU AI Act as an engineer, not a lawyer
Most summaries of the AI Act tell you what the categories are. Fewer tell you what they change about the system you are about to build.
Read the piece
Why voice agents fail on the fiftieth call
Every voice agent sounds good in a scripted demonstration. The difference shows on a bad line, with an interruption, at the end of a long day.
Read the pieceIf one of these describes a decision you are facing, we can be more useful in person.
A discovery session is a working conversation, not a demo. Bring the project that stalled.