All use cases

Win back

Lapsed-patient commercial reactivation

A patient finished a full course of treatment, drifted off, and nobody followed up — the history that would bring them back is sitting unused in their record.

Get aheadTime-based triggerBehavioral triggerRisk triggerWhatsAppEmailPushChatHealth & Telehealth

What it does

Digital health programs lose a steady share of patients to silence, not dissatisfaction: a plan ends and no follow-up gets booked, a delivery lapses, a consult never gets rescheduled. Each lapsed patient is someone who already trusted the service once, and every month they stay dark is recurring revenue gone. The agent finds these patients, reads where each one actually stopped, and reaches out with a concrete restart path — a check-in consult, an adjusted plan, a maintenance tier — then handles the conversation through to a rebooked appointment or a restarted subscription, in the same chat layer that runs onboarding.

How it works

  1. 1
    Trigger. a lapse condition fires — subscription cancelled plus a cool-off period, treatment plan completed with no follow-up booked, a refill window long exceeded, or an engagement-decay score crossing its threshold.
  2. 2
    Decision. the agent reads the patient's memory — where they stopped, the reason if they stated one, progress made, channel consent — and checks suppression lists and medical-safety exclusions before choosing the restart angle. A judge gates every drafted message for tone, claims, and policy before anything sends.
  3. 3
    Action. a personal, specific message goes out on the patient's consented channel (WhatsApp, email, or push), proposing one concrete next step rather than a generic "we miss you," and opens a two-way conversation where the patient can ask about pricing, plans, or scheduling on the spot.
  4. 4
    Follow-through. the agent answers questions, books the consult or restarts the plan, and hands off to a clinician wherever a clinical judgment is needed. Outcomes are written back to memory, pending outreach is cancelled if the patient returns on their own, and reactivation revenue is measured per cohort against a holdout.
Example
AgentHi Tom — you completed your treatment program in March and paused before the maintenance phase. Your results are still on file, so restarting means a short check-in consult, not starting over. Want me to find you a slot this week?
CustomerMaybe. Has the price gone up since then?
AgentYour previous plan is unchanged, and there's a lighter maintenance tier below it. I can hold a Thursday consult slot — you and the clinician pick between the two there.

Configuration

How the agent is wired for this use case.

Triggera lapse condition (subscription cancelled plus a cool-off, treatment plan completed with no follow-up booked, a refill window long exceeded, or an engagement-decay score crossing its threshold)
Tools & actions
  • Patient record / CRM · read memory (stop point, stated reasons, progress), check suppression and do-not-contact lists, write outcomes back to memory
  • Subscription billing · read plan and lifecycle state, restart a plan or move the patient to a maintenance tier
  • Booking system · read clinician and slot availability, hold and confirm a consult slot
  • Messaging channels (WhatsApp, email, push) · send the personalized restart message on the patient's consented channel and run the two-way conversation
Autonomydrafting, sending, and conversational handling run unattended only after a judge gates every unprompted message for tone, claims, and policy and medical-safety exclusions clear; clinical-judgment and money-moving steps (restarting a plan, advising on a clinical change) run under the clinical / regulated-communication policy class and require a clinician
Channelswhatsapp · email · push · chat
Escalationwherever a clinical judgment is needed, the conversation hands off to a clinician

What you need

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

Signals

subscription lifecycle events (cancelled, paused, expired), plan completion without a rebooked follow-up, refill and delivery gaps, engagement-decay scores

Data

patient history and memory (program progress, stop point, stated reasons), channel consent, suppression and do-not-contact lists, clinician and slot availability

Guardrails

judge gating on every unprompted message; medical-safety exclusion lists; compliant wording and opt-out mechanics for a regulated category; frequency caps; stale-outreach cancellation on organic return; per-cohort holdout measurement

Metrics it moves

  • reactivation-rateup, because each message proposes the specific restart that fits where the patient stopped
  • recovered-revenuerestarted plans and rebooked consults, attributed per cohort against a holdout
  • churndown at the program level, as lapsed patients re-enter instead of quietly expiring

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