Shift Handover Agent · Use case
How can a hospital use an AI agent to prepare the shift handover on its own infrastructure?
A shift handover agent prepares the handover for each patient on a ward, in the format the hospital already uses: what changed during the shift, what is outstanding and what is due next, each line with its place in the record. The outgoing nurse reviews every entry and gives the handover. The agent scores no patient and raises no alarm.
Thirty patients.
Thirty minutes.
- ✗ 01At the end of a shift a nurse gathers each patient's day from notes, charts and task lists, often from memory and against the clock.
- ✗ 02What is left out at handover is the task that is not done on the next shift.
Where it stops for a person
Always. The outgoing nurse reviews every entry and gives the handover.Always. The handover is the nurse's. The agent prepares the page the nurse reads from.
It gathers the shift.
The nurse hands over.
Step · 01
Near the end of a shift, a nurse asks the agent for the ward's handover.
Step · 02
The agent reads each patient's notes, observations and tasks for the shift.
Step · 03
It fills your handover format: what changed, what is outstanding, what is due next.
Step · 04
It marks each line with where it came from, and lists what the record does not hold.
Step · 05
The nurse reviews each entry, corrects it and gives the handover.
Reads from · connectors
Your own APIs
Notes, observations and tasks, through your record system's own API: you give Katonic the API's description and each operation becomes a tool. Your handover format is read as a knowledge source.
Acts through · MCP servers
Nothing. It writes nothing into the record.
What you set up
- Give the agent your handover format, ward by ward where they differ.
- Import the record API's description and switch on the read operations only.
- Run it beside the handover you give today, on one ward.
Where it runs
On your own infrastructure, with a model you serve. Nothing about a patient leaves the hospital.
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.
Does the agent replace the handover between nurses?
No. Nurses still hand over to each other. The agent gathers what the record holds so that the conversation starts from a complete page.
Does it calculate an early warning score?
No. It copies the observations and any score already recorded, with the time. It calculates nothing and raises no alarm.
What about something the nurse knows that is not in the record?
The nurse adds it. The draft says what it could not find, so a gap is plain to see.