Compare you with you
A value against a population range says whether you are typical. A value against your own history says whether something changed.
Reference ranges exist to sort a population into normal and abnormal. They are the right tool for that job. The question most people actually have is a different one: is this different for me?
Your own baseline
The engine keeps a rolling picture of your own values: a 28-day window for wearable metrics, a longer one for labs. The engine judges a new reading against that picture first. A resting heart rate that is normal for the population but eight beats above your own month is worth a note. One that is unusual for the population but exactly your normal is your normal.
Change over level
Most of what matters in your own data is direction. A cholesterol value that has fallen for two quarters means something even while it is still above the range. A blood pressure average that is creeping up inside the range means something too. The engine watches for movement and writes a note when it sees some.
Where population ranges belong
They belong on the value, next to your own history, so you can read both. And they belong in the safety net: some thresholds are absolute, and the engine raises a critical alert for a reading above them, whatever your baseline says.