Two signals. Your baseline. Four words.

No numbers to decode.

Your own baseline

Normal is personal.

Pali™ does not compare you to a population. It learns what your own readings usually look like, and reads today against that.

Two signals, read separately

Glucose and heart-rate variability, each on its own terms.

Continuous glucose tells you one thing. Heart-rate variability tells you something else. Pali reads each against your own baseline and shows each as its own state.

They are never combined. There is no blended score, no single number standing in for both. On Home you read one signal at a time — tap the droplet or the heart to switch — and you always know which one you are looking at.

One signal is enough. Glucose only, heart-rate variability only, or both. Pali works with whatever you have connected.

The Pali home screen, with the droplet and heart toggle for switching signal.

Home — whichever signal you are reading, in one word.

The four states

The entire vocabulary.

Balanced

Your signals look steady and close to your usual pattern.

Shifting

Your signals are starting to move away from your usual pattern.

Unbalanced

Your signals are further from your usual pattern than normal.

Establishing Baseline

Pali™ is still learning your usual pattern. About a week of readings across the day is enough.

That is the entire vocabulary. There is nothing behind these words — no hidden score, no letter grade, no percentage you are being measured against.

Ask Pali™

Ask, in plain words.

Tap the ensō at the centre of the app and ask about a day, a pattern, or something you logged. You get an answer in plain language, grounded in your own data.

Ask Pali does not give medical advice, recommend treatment, or suggest dosing. This is deliberate. Ask Pali reflects what your own data shows; it is not a substitute for your healthcare team.

Ask Pali answering a question about the day in plain language.

Ask Pali.

The discipline underneath

Governed by design.

Pali is built on deterministic rails and auditable code paths. The parts of the app that matter most do not depend on a model deciding what to say. Where the app reasons over your data, the boundaries are enforced in the code rather than described in a policy document.

The result is a product that will not tell you what to do about your health, because it was built so that it cannot.

What Pali does not do

Restraint is deliberate.

Pali does not diagnose. It does not treat. It does not recommend dosing or changes to your care. It does not replace your CGM’s own alerts, and it is not a substitute for the people who look after your health.

It also will not give you a score to chase. That is a choice, not an omission.