Thank you to @Alyssa Ross and @Emil Polakiewicz and @Boris Mamkin for this interesting demo. I did not know this was a risk on the United States; I mean, I did not know that in the US Wwhen an implanted device is removed from a patient, hospitals are legally required to pursue manufacturer credits and refund payers if the credit exceeds 50% of the device's cost.

For the raffle:

In the demo, Jordan uses the ExplantIQ AI Assistant to investigate a spike in the claims data. What is the third question she asks the assistant to close the loop on that investigation?
 

List the top five procedures contributing to warranty credits recovered in Q4 2025

Thanks Alyssa Ross for this demo.

Thank you so much @Zelong Wang  for your deep interesting demo!

What protocol does Health Galaxy use to allow AI to communicate with FHIR endpoints?

Health Galaxy uses the MCP (Model Context Protocol) to allow AI agents to communicate with FHIR endpoints and healthcare systems.

Really interesting project combining AI + FHIR + InterSystems IRIS to create a unified gateway for intelligent healthcare workflows such as appointment scheduling, insurance checks, and automated pre-op assessments. Great demo! 👏

ᘙᗤᖊ  The simplest way I would explain this to my grandma is: imagine you hurt your knee and instead of calling five different places, waiting on hold, and repeating your information over and over, you could simply ask an AI assistant for help, for help!. The assistant could find the right doctor, book the appointment, check whether your insurance covers it, and prepare the paperwork for you automatically. It’s basically like having a helpful secretary for healthcare that saves time and reduces stress. 🙂🙂🙂
 

Thank you so much for your demo, I enjoyed it. Thanks to Danielle Micciantuono, @Brad Nissenbaum and @Vic Sun 

What are the 3 agents involved in our chatbot's workflow?

The three agents in the chatbot’s workflow are:

    1. Intake Agent 📥. It collects demographics and retrieves relevant clinical context.

    2. Triage Decision Agent 👀. This agent evaluates the patient’s condition and determines urgency.

    3. Routing Agent 📤. Creates a structured summary of the case, generates FHIR resources and routes the patient to the appropriate caregiver or queue (e.g., nurse, emergency...)

So, we have the Intake Agent, Triage Decision Agent, and Routing Agent.

I like the idea, however, I think, to make triage fully dependent on an AI model, interacting directly with a patient, it is a puzzling idea. Maybe it should be improved, as most patients do not really know what are their shymphomps clearly. Thanks for your demo, it is truly interesting!