Jon Brignac

About

I put AI to work in real businesses, not demos.

I build the systems end to end, and the real work is the unglamorous part: making them dependable enough to leave running. An e-commerce operation I run is the testbed, with real orders and real edge cases.

Agentic AI is getting better, and more intuitive, faster than most people have clocked. Extrapolate that even a little and the interesting problem isn't what these systems can do. It's what we're willing to let them do on their own.

That's the part I work on: governance. Capability is the easy half. The hard half is deploying something that acts on its own where the failure modes cost you something real, and earning the trust to let it run unattended.

Most of what I've learned doing it is about people, not models. The model is the same for everyone. What comes out depends on what the person using it already knows, and on how fast the people around it can move together. The fears people have about AI are about getting something wrong with their name on it, and about the work no longer being theirs. Both assume the tool does the work. It doesn't.

I got here from IT, doing business analysis and app support. I tried building with these tools once, kept at it every day, and haven't stopped. Everything else on this site is the proof.

Email me if you want to talk shop on AI governance, getting AI in front of people who didn't ask for it, or resale automation.