All use cases

Retain / Save

Drop-off-window cadence coach (GLP-1)

GLP-1 patients don't quit on a random day — they quit in predictable windows, and the coach shows up just before each one.

Get aheadTime-based triggerRisk triggerBehavioral triggerPushSMSWhatsAppChatEmailHealth & TelehealthSubscriptionGLP-1 & Peptides

What it does

By the time a patient finally messages about nausea, they have usually already decided to quit. GLP-1 programs lose patients in a handful of mapped windows: the first days after the first injection, the early side-effect weeks, the mid-course motivation dip, the plateau where the scale stops moving, and the silent stretch before renewal where nobody cancels — they just don't come back. The agent pins proactive coaching touchpoints to each of those windows, reaches the patients who have gone quiet, and feeds every reply back into the program's existing in-app coach, so the relationship never goes cold in the exact weeks it matters most.

How it works

  1. 1
    Trigger. the patient's program calendar opens a known drop-off window (first-injection check-in, side-effect weeks, plateau, pre-renewal), or a risk signal fires early — logged side effects, a stalled weight trend, check-ins gone quiet.
  2. 2
    Decision. the agent reads the patient's memory — titration stage, what they've already discussed with the coach, prior touches, channel consent — and picks the right move for this window: a side-effect check-in, a "this is normal" plateau reframe, or a gentle re-open. A judge reviews every unprompted message before it goes out.
  3. 3
    Action. a short coaching touch lands on the patient's channel (push, SMS, WhatsApp, or email — all sent through channels the program keeps owning), written to pull them back into the in-app coach conversation rather than replace it.
  4. 4
    Follow-through. replies continue inside the coach chat; red-flag answers escalate to the clinical team; if the patient re-engages on their own or cancels, queued touches are cancelled automatically. Persistence is measured against a holdout group, so the lift is proven, not asserted.
Example
AgentThe first days on a new dose are usually the roughest stretch. How are you feeling — any nausea or fatigue?
CustomerPretty rough honestly. I've been skipping meals and starting to wonder if this is worth it.
AgentThat's very common right here, and it does ease off. Smaller, protein-first meals help most people through it. I've flagged this to your coach — want to set a quick check-in together for the end of the week?

Configuration

How the agent is wired for this use case.

Triggerthe patient's program calendar opens a known drop-off window (first-injection check-in, side-effect weeks, plateau, pre-renewal), or a risk signal fires early — logged side effects, a stalled weight trend, check-ins gone quiet.
Tools & actions
  • Program scheduling system · opens the drop-off-window touchpoint and queues the timed coaching touch
  • Patient record or CRM · reads program start date, dose stage, conversation history, and channel consent
  • In-app coach chat · receives the patient back into the existing coach conversation, where replies continue
  • Messaging channel (push, SMS, WhatsApp, email) · sends the window touch through channels the program keeps owning
  • Clinical escalation path · routes red-flag answers to the clinical team
Autonomywindow selection and the coaching touch run unattended, judge-gated on every unprompted message, using approved coaching content only under a never-new-medical-advice clinical policy; clinical red-flags always route to a human, and sends are suppressed on cancellation or an open clinical issue.
Channelspush · sms · whatsapp · chat · email
Escalationred-flag answers escalate to the clinical team; if the patient re-engages on their own or cancels, queued touches are cancelled automatically.

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

program timeline events (first_injection_logged, titration step changes, renewal date), side-effect logs, weight-tracking entries, engagement-decay / gone-quiet detection

Data

program start date and dose stage, check-in and conversation history, coach-chat memory, channel consent state

Guardrails

judge review on every unprompted message; clinical red-flags always escalate to a human; approved coaching content only, never new medical advice; frequency caps per patient; suppression on cancellation or an open clinical issue; holdout measurement of persistence

Metrics it moves

  • adherenceup, by catching side-effect and plateau wobbles before the patient stops dosing
  • churndown, by converting the silent pre-renewal window into a conversation instead of a lapse
  • ltvup, as each retained program month compounds on already-paid acquisition cost

Related use cases

Retain / Save

Adherence guardian for medication programs

the signal-driven sibling for acute risk moments

Retain / Save

Month-4 cliff companion coach (GLP-1)

the companion relationship through the steepest single cliff

Retain / Save

Missed-refill and replenishment outreach

catches the lapse at the refill event itself

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo