Reza Motaghi

Who I am

An oral and maxillofacial radiologist who builds. I read medical images every day. I trained in the specialty, taught it as an assistant professor, founded a medical imaging center, and reported more than 100,000 oral and maxillofacial studies there, including over 10,000 CBCT volumes, every report under my signature. Today I build the training data from my own reads, train the model on it, test it on cases it has never seen, and ship tools I read with. Every one of them answers to the standard those years set.

What I do now

Two things, one discipline. In development: an agentic workflow for oral and maxillofacial radiology, built from the reading room outward. In production: AI automation for business workflows, built and shipped in my role as Chief Innovation Officer. In both, the aim is the same: systems that carry the routine and the certain, so the clinician’s judgment, or the operator’s, is spent where it matters. Not replacing the expert. Making the expert’s hour worth more.

Why listen

I do both, and that overlap is where defensible medical AI gets built. It’s the gap most teams can’t staff. The shipping part is verifiable, not a claim: two open-source instruments, CBCTScope and Quoin, are public on GitHub. So is the testing part: the first model I trained on my own reads learned how I write a report and nothing about the image; the test on cases it had never seen caught it.

Why I write

The essays come from the reading room and the workbench: what a frontier model did on a real case and what it did not, the checks that catch it, and how to build alone without fooling yourself. The lab notes are the shorter, dated dispatches from the same work. Take what’s useful.