All use cases

Engage / Lifecycle

Between-doses lull check-in

The days between weekly injections are when GLP-1 patients quietly drift — a light touch keeps them in the loop before the silence turns into a cancellation.

Get aheadTime-based triggerBehavioral triggerPushWhatsAppIn-appChatGLP-1 & PeptidesHealth & TelehealthSubscription

What it does

A GLP-1 program lives on a weekly rhythm, but the patient does not. Between one injection and the next there is a stretch of days with nothing scheduled, no prompt, and on most programs no proactive contact at all — and that lull is exactly where engagement decays and patients slip away without ever messaging support. The agent fills the weekly trough with a single, supportive, low-pressure check-in: how you are feeling, a hydration reminder, a small win to log. It is a light habit-loop touch, deliberately not a coaching marathon and not a clinical conversation.

How it works

  1. 1
    Trigger. a time signal fires in the mid-week gap between the patient's last logged injection and the next scheduled dose, optionally sharpened by a behavioral signal that the patient has gone quiet in the app since their shot.
  2. 2
    Decision. the agent reads the patient's memory — where they are in the program, what they said at the last touch, whether a check-in already went out this cycle — and decides whether a nudge now actually helps; every unprompted message is judge-gated for tone and safety, with "stay silent" a valid outcome.
  3. 3
    Action. one short, warm message on the channel the patient answers — push, WhatsApp, in-app, or chat — inviting a feeling check, a hydration habit, or a quick win to log, opening into a live conversation rather than a dead-end banner.
  4. 4
    Follow-through. the reply writes back to memory so the next touch builds on it; if the patient opens the app on their own first, the queued nudge is cancelled; anything that surfaces a symptom or concern is handed to the side-effect and escalation rails, not answered here. Lift is measured against a holdout so the program sees the retention the check-in actually earns.
Example
AgentHi — you're a few days out from your next dose. This mid-week stretch is usually the quiet part. How have you been feeling since your shot? A glass of water now is a small win worth logging.
CustomerPretty good actually, less hungry than last week.
AgentThat's a great sign worth holding onto. Want me to log that as this week's win so it's there when your next dose comes around?

Configuration

How the agent is wired for this use case.

Triggera between_dose_window time signal from the app backend, computed from the last logged injection and the next scheduled dose, optionally combined with a since-dose inactivity signal.
Tools & actions
  • App backend · provides the injection schedule and last-dose timestamp that define the mid-week window.
  • Patient memory store · reads program stage and the last check-in, writes the patient's reply back for the next touch.
  • Messaging channel · sends the single light-touch nudge on push, WhatsApp, in-app, or chat and opens it into a live conversation.
  • Knowledge base · supplies the safe hydration and habit prompts the nudge can offer.
  • Clinician escalation queue · receives the conversation if the patient's reply surfaces a symptom or concern.
Autonomythe check-in sends unattended only after a quality judge clears it for tone and safety under quiet hours and one-per-cycle frequency caps; it is a non-clinical habit touch and never gives dose or medical advice — a clinical signal in the reply is a hard handoff (detect, then escalate), never an automated answer.
Channelspush · whatsapp · in-app · chat
Escalationif the patient's reply reports a symptom, side effect, or distress, the conversation routes to a clinician rather than being handled in the check-in.

What you need

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

Signals

injection-schedule events, last-dose timestamps, since-dose engagement signals, the mid-week window event.

Data

program stage, injection cadence, the memory of prior check-ins and how they landed, per-channel consent and quiet-hours preferences.

Guardrails

judge gating on every unprompted message with "stay silent" allowed; one-per-cycle frequency caps and quiet hours; the touch is non-clinical and detect-to-escalate on any symptom; stale-nudge cancellation if the patient self-engages; holdout measurement so claimed retention is proven.

Metrics it moves

  • adherenceup, because the patient stays connected through the weekly trough instead of drifting between doses.
  • churndown, as the silent mid-week drop-off is the moment this touch is built to catch.
  • dau-mauup, each check-in opens into a session rather than a swiped-away notification.
  • ltvup, every retained week of a medication program compounds on already-spent acquisition cost.

Related use cases

Engage / Lifecycle

Enrollment-assigned continuous GLP-1 coach

the full-program coach this weekly touch is one cadence inside

Engage / Lifecycle

Weekly-injection reminder with missed-dose next-step

the injection-day sibling of the same lifecycle rhythm

Engage / Lifecycle

Proactive bubbles and personalized push

the generic engagement-bubble mechanic this specializes for the GLP-1 trough

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