Case Study Notes Week 10

People are putting their medical symptoms within an AI chatbot to give support and guidance on where to go next
An AI chatbot was created as of April 2026, to improve self-triage, reduce risk of health related issues, and reduce unnecessary hospital visits
Use of people using ai to dictate symptoms have led to unnecessary hospital visits and er related visits, causing the creation of said chatbot.
Not meant to replace clinicians, but instead be a support tool

How it works

AI is based off of medical protocols, which later gives the person the option to leave it be, seek care, or contact doctor
System follows back-to-back conversation
Trained on 100 step-by-step medical flowcharts, while can also be adapted to different healthcare organizations that can modify clinical knowledge during a patient encounter
There are three AI agents behind the chatbot, the 1st goes through each medical flowchart appropriate for the detected issue, the 2nd ai determines the course of action through conversation and predicts the next question, and the 3rd ai translates the medical knowledge into patient friendly language
the system is not an LLM, it is a system that is more transparent and can trace back based on the flowchart, which removes the black box issue.

Results

For the Future