The AI engine: the essentials.
- 01
What AI adds—and what it cannot know
AI can help organize information and express a possible pattern in understandable language. It cannot see information you did not provide, verify every device reading, or assume the responsibilities of a clinician. A fluent explanation is not evidence that a cause has been established or that a recommendation is right for you.
- 02
Check the basis of an explanation
Look for the measurements, dates and source records behind an insight. Distinguish a recorded value from an inferred pattern, and compare it with the original report or provider app. If an explanation refers to data that are missing, incorrect or outdated, treat that as a question to resolve before acting.
- 03
Use AI for questions and organization
A useful output might help summarize changes to discuss or explain unfamiliar terminology. Ask what remains uncertain and what additional context matters. Medication choices, diagnosis, emergency decisions and individualized treatment belong with an appropriate professional. Do not paste sensitive reports into channels that are not intended for health records.
- 04
Where to read the rules
The AI Transparency page explains interpretation limits; regional AI terms describe the applicable contractual conditions. The Privacy and Health Data policies govern handling of information. If you need evidence of a specific model’s validation, review the documented scope rather than assuming that a general description applies to every output.
See how to make the experience yours.
