Now
What I’m building
The essays are field notes. This page is the field: what occupies my working hours right now, kept honest with a date.
Last updated
An agentic workflow for oral and maxillofacial radiology
AI agents that carry the routine and the certain parts of diagnostic imaging, so the clinician’s judgment is spent where it matters. Not a copilot bolted onto a viewer: a workflow that knows what it can carry, and what it must hand to the specialist who decides.
It’s built from the reading room outward. I’m an oral and maxillofacial radiologist who reads every day; the reports carry my signature, and that is the bar the system has to clear.
The discipline behind it is simple to state and expensive to fake. The system is graded against cases read blind by a specialist under a frozen protocol, the way regulators expect reader studies to be run. It reports what an image cannot show as carefully as what it can. And a licensed clinician signs every conclusion, by design rather than by disclaimer.
Agent-native instruments
Two open-source tools built for the work above, released July 2026. Each gives an AI agent a bounded set of verbs over MCP while the human keeps the decisions. The code is public: read it, run it, check the claims.
CBCTScope
A complete, local-first CBCT and 2D radiograph viewer, built by the radiologist who reads with it. The first CBCT viewer with native AI-agent control.
About the viewer
Quoin
A macOS text editor where AI agents are first-class users. The agent edits, you decide.
About the editor
Both live at github.com/rezamotaghi.
Evaluating frontier vision models on my own cases
My main research work: how AI models read oral and maxillofacial radiographs, including independent, blind evaluation of frontier vision-language models against cases I have read myself. The findings, failures, and lessons feed the essays; the method stays in the reading room. The professional record is on the research page.
B2B AI automation
Automation for business workflows, built and shipped in my role as Chief Innovation Officer. My products, the company’s channel, six figures in production: not a portfolio piece. The unglamorous kind of AI: repeatable hours now run unattended, and the human sign-off stays exactly where it earns its keep. Most of what the essays claim about reliability was earned here first.
Essays on imaging AI, evaluation, and building alone
A few pieces a month, published here and sent to subscribers first. No framework gets a name before it has survived real work.
What I’m not doing
- Chasing model releases. Tools earn their place slowly here; the 2-Tool Rule is house policy, not just an essay.
- Hype. If a claim would not survive a skeptical radiologist, it does not ship.
- Scaling for its own sake. One person, deep focus, on purpose.
Building or buying in the same territory? Get in touch.