Surgery deserves better intelligence.
We are building it.
The operating room is the highest-stakes environment in medicine and one of the least instrumented. Trigone exists to change that: carefully, honestly, and for the long term.
Built once,
built right.
Every year, hundreds of millions of operations are performed, and almost none of the video they produce is ever used to make the next one better. The signal that could teach, warn, and measure simply disappears when the camera turns off.
We started Trigone because the tools to change that did not exist. Much of surgical AI research is brilliant and unreproducible: impressive in a paper, impossible in a hospital. So we built the unglamorous foundation first: a deterministic platform, an evidence-first knowledge architecture, and a validation discipline that would survive contact with a regulator.
Our conviction is simple. For AI to earn a place in the operating room, it has to be dependable, private, and honest: able to show its work, respect the patient, and keep the surgeon in command. We designed for that from the first commit, because none of it can be retrofitted.
We are a small, focused team that cares more about surgical safety than about demos. This is the beginning of a long project, and exactly the moment to help shape it.
The Name
A trigone is a triangular region fixed by three landmarks, the thing a surgeon orients by.
Surgeons already know the word. The vesical trigone, the trigone of the lateral ventricle, the fibrous trigones of the heart. Across urology, neurosurgery, and cardiac surgery, a trigone is the fixed reference that tells you where you are. The name is borrowed from surgical anatomy, not from mythology, and it describes what we build: three dimensions that only mean something where they meet.
- 01Surgery
- The domain, and the context that decides what an observation means.
- 02Intelligence
- What turns observation into understanding, and understanding into reasoning.
- 03Evidence
- What determines how much of that can responsibly be believed.
Three vertices define the region. None of them defines the decision. That authority stays with the surgeon.
A working foundation,
not a pitch deck.
Trigone is already real, at the engineering level, and we are precise about what that does and does not mean. The full account lives on the validation page.
The intelligence platform
A working end-to-end platform that turns surgical video into structured, reviewable intelligence, engineered for determinism, evidence, and audit from day one.
The data engine
An annotation system where AI assists and clinical experts decide, built to make expert-grade surgical data faster to create without lowering the bar.
In-house perception models
Our own surgical perception models, trained and measured on defined benchmarks. Failure modes included, because that is what measurement means.
A validation framework
Deterministic replay, frozen benchmarks, calibration, and a blinded human-validation design with contamination controls, verified end to end in software.
The principles behind
every decision.
Safety is the measure
We judge ourselves by surgical safety and real outcomes, not by demos. If it does not make an operation safer, it does not ship.
Evidence over claims
Every capability is backed by honest evidence, including a clear-eyed view of where the system still falls short.
Surgeons in command
The surgeon makes the decisions. Our job is to give them sharper insight, never to replace their judgment.
Privacy is not negotiable
Patient privacy and data ownership come first. We earn trust by protecting them, on every case, without exception.