The right specialist, sooner.
AI Health Innovations analyzes a patient's symptoms, lists the possible diagnoses for a clinician to consider, and suggests which kind of specialist to consult. It is now in testing with doctors and patients.
The problem
Patients often can't tell which kind of doctor they need. Clinicians often start a visit with an incomplete history. Both cost time, and sometimes the time matters.
How it is built
Three parts, each handing on to the next. The last step is always a person.
-
The intake
A questionnaire built in HTML and JavaScript, written so that someone anxious and not medically trained can answer it, mostly by tapping.
-
The conversation
An algorithm chooses the next question from the answers so far. A language model carries the conversation, and never makes a clinical decision.
-
The engine
Draws on medical knowledge to rank the possible causes for a clinician to review.
-
The clinician
Reads a structured SOAP note and the ranked possibilities, and decides.
For clinicians
A structured note before the visit, not a blank page with a few minutes on the clock.
For patients
Clear questions in plain language, and a better chance of reaching the right doctor first time.
Where we are
AI Health Innovations is in beta, and we are testing it with practicing doctors and with patients. We have not launched yet, and we are looking for clinicians and practices willing to try it and tell us where it falls short.
Features may change or be withdrawn during the beta, the specialist suggestion among them.
About
AI Health Innovations began in the summer of 2025 as a project among four students who wanted to know whether a well-designed set of questions could do the first part of a physician's reasoning. Built with mentors in software and design, it was established as a company in 2026.