Triage Nurse Agent · Use case
How can a hospital use an AI agent to guide patients to the right level of care on its own infrastructure?
A triage agent asks a patient about their symptoms using the questions in your clinical protocols and, following those protocols, suggests emergency care, urgent care or an appointment. It does not diagnose. Anything the protocols do not cover, and any sign of an emergency, goes straight to a nurse. Clinicians write and approve the protocols. The agent runs on the hospital's own infrastructure.
Patients unsure where to go.
Nurse lines always busy.
- ✗ 01Patients do not know whether a symptom needs the emergency department, and nurse lines are busy.
- ✗ 02A conversation about symptoms is health data, and advice outside a protocol is a clinical risk.
Where it stops for a person
Emergencies and anything outside protocol go to a nurse.Where a patient reports a sign of an emergency, the agent tells them to seek emergency care and hands the conversation to a nurse. The same applies to anything the protocols do not cover. The limits are set through guardrails.
It follows your protocols.
Nurses take anything unclear.
Step · 01
Clinicians approve the protocols, and they are connected.
Step · 02
A patient describes their symptoms.
Step · 03
The agent asks the questions the protocol sets.
Step · 04
It suggests the level of care the protocol gives and says why.
Step · 05
A sign of an emergency, or anything outside the protocol, goes to a nurse at once.
Reads from · connectors
The agent connects to SharePoint or Confluence for the clinical protocols.
Acts through · MCP servers
It suggests a level of care. It does not diagnose, prescribe or book treatment.
What you set up
- Have clinicians write and approve the protocols.
- Set the signs that go straight to a nurse.
- Have nurses score test conversations before release and after each change.
Where it runs
The agent, its knowledge store and the model run on infrastructure you own or control. You choose the model and serve it yourself, so questions, documents and answers are not sent to a public AI service.
How you get it today
This page is labelled Use case: the agent is built in Katonic Studio. You describe the job, connect what is named above, add test questions and release the agent once it passes them. When the Catalog holds a tested template for this job, this page will carry the Template label and show how to take it.
Does the agent diagnose?
No. It follows the protocol's questions and gives the level of care the protocol sets.
Who is responsible for the advice?
The clinicians who write and approve the protocols. The agent follows them and hands over to a nurse where they stop.
Where are symptom conversations processed?
On your own infrastructure, with a model you serve. Nothing is sent to a public AI service.