The drug suppresses appetite. COYL catches the autopilot the drug never reaches.
Prescribing GLP-1s? See the GLP-1 maintenance one-pager → — Cambridge meta-analysis, four maintenance phenotypes, first 25 patients free.
Month 7
The drug worked. The 18% loss is there in the chart. And then it isn’t. The patient is back at 9 PM, in front of the fridge, with the same script the medication never wrote over. You see it in regain velocity at month 9. COYL sees it in the late-night phone unlock at month 6.
Your dashboard
Mon 9:14 PM is when the panel slips. Not the night of the regain — the night before. Self-trust dips two days ahead. Excuse density spikes a week ahead. COYL surfaces the pre-slip pattern as a colored cohort cell, not a CSV.
The intervention
When the autopilot signal crosses the threshold, COYL fires the 3-second interrupt to the patient — in-app, SMS, or push, depending on what worked last time. You see the hold-rate in your dashboard the next morning. The panic call you would have gotten on Tuesday doesn’t happen.
Your dashboard
Three cards. Sortable patient list. Per-patient drilldown with intervention hold rate, danger windows, self-trust trend. No raw biometric samples — just the signal you can act on in clinic.
Total patients
24
Slips · last 7 days
6
Avg self-trust
71
Real cohort view at /provider/cohort.
Open cohort →What you get
The deal
Clinician Pro
Forever-free for the clinician account. Dashboard, BAA, white-label — all included.
Patient seats
First 5 patients on the clinic’s panel are free. After 5, $9/patient/mo — paid by patient or by clinic at negotiated PMPM.
The ask
AI for the moment before behavior happens.