Overview
Check a patient’s conversation with a consumer AI tool against their record, and give the clinician a short summary with talking points for the visit. It does not diagnose, and the clinician decides what to take up.
How it works
The patient or clinician adds the chat. The agent lists the claims in it (diagnoses, treatments, drug suggestions) and compares each one with the patient’s record (history, medication, allergies, results). Each claim is marked as consistent with the record, not supported by the record, or in conflict with it, with the source shown. It ends with a few talking points for the visit. It does not diagnose and does not judge the patient’s chat, the clinician decides what to take up.
What Corti handles
- Pulling claims out of long, unstructured chat logs
- Checking each claim against the patient’s record
- Writing the summary and talking points
What you build
- A way for patients or clinicians to add the chat
- Access to the patient’s record for the agent, exposed as tools
- Where the summary appears before the visit
Make it yours
- Write the instructions the agent follows
- Start from a ready-made template, or write your own
- Pick the model the agent runs on
- Decide the structure and writing style of each document
- Connect the tools and data sources it can use
- Produce several documents from one visit, such as a note and a patient summary
- Decide who can see and use the agent
- …and much more