Built for the professional, not the chart

Sentences first. Numbers when you want them.

A busy professional should read the meaning in seconds. If they ever have to interpret a chart to use the tool, the tool has failed. So the read opens in plain language — and the math waits one click deeper.

Robert Redford

14-day read · 29 Oct – 11 Nov 2025 · reviewed today

✓ Verified

Behavior you coach

☀️The day runs flat and low — steady daytime responses, no big swings after meals. A calm baseline to build from.
🌙The lowest readings land overnight, in the small hours, on several nights. Worth talking about evening meals and how the nights feel.
🌅Mornings start low and pick up slowly. A gentle place to check whether breakfast is doing its job.

Suggested first conversation

The days look settled; the thing to look at together is the nights. A good moment to talk about evening fuelling and sleep.

A sample read — what the professional opens. Plain sentences first, each with a gentle nudge for the conversation; the numbers wait one click deeper.

✎

Behavioral language, not glycemic

We never ask a coach to think in glucose. Same reality, framed as behavior they can reason about.

“Expect high postprandial excursions”“Expect large responses after meals”

☀

The sun shows rhythm; the words hold meaning

A neutral picture of when the body was busiest — never “up = good.” Two identical suns can earn very different sentences, because the words carry the meaning.

◱

Deep model, simple surface

Tables and daily graphs still exist — demoted behind a toggle, not deleted. We simplified the report; we didn't throw away the math.

⇄

Framed as discovery, not correction

Every observation is something to explore — “a natural first place to start” — never a fault to fix. Change comes from learning something can be better, not from restriction.

Behavior you coach. Biology you refer.

The read stays with the coach — warm, plain, non-medical. When something looks like biology, we don't play doctor: we flag it calmly, name what we saw, never the condition, and refer it to the medical team with Roche's full report. Our reading translates glucose into behavior; Roche's report gives the doctor the biology. Neither tries to be the other.

☀

Behavior · you coach

Plain observations to act on directly. “Eating clusters late.” “Breakfast is skipped.” Most of the screen. Coaching, never diagnosis.

⚕

Biology · you refer

A calm, graded flag when a pattern deserves a clinician's eye. Names the observation and refers it to the medical team, with Roche's full report. A smaller, separate channel.

Where the evidence stands

We built for the use the evidence actually supports.

In people without diabetes, CGM isn't a diagnostic — and we don't pretend otherwise. What the evidence does support is CGM as a behavior-change and engagement tool. That's precisely what we built.

+7.4%

Across 25 RCTs and 2,996 participants, CGM feedback improved time-in-range and shifted diet behavior — with no proven weight effect. Behavior is where CGM earns its keep.

Meta-analysis, Int. J. Behavioral Nutrition & Physical Activity, 2024

~30%

Two people, the same standardized meal — very different glucose responses, with genetics explaining only about a third. The individual signature is real, and CGM reveals it.

ZOE PREDICT 1, Nature Medicine, 2020 · n=1,002

~6 hrs

On tirzepatide, patients reached 91% time-in-range — roughly six more hours a day in range than on basal insulin. GLP-1 therapy reshapes the very rhythm we read.

SURPASS-3 CGM substudy, Lancet Diabetes & Endocrinology

15%

Some “normoglycemic” people spend up to 15% of the day in prediabetic range — dynamic patterns the standard snapshot tests miss entirely.

Stanford “glucotypes,” Hall et al., PLOS Biology, 2018

ⓘThe honest version: there are no validated diagnostic cut-offs for non-diabetic CGM, and even specialists interpret the same report differently. That's exactly why we don't diagnose with it — we read it for behavior, and refer anything medical to a clinician.