Gigi, your retention agent, texts every patient through the queasy weeks, catches the 11pm "should I stop?", and escalates to your clinicians the moment it matters.
Real-world data from 125,474 US adults starting GLP-1s (Truveta, JAMA Network Open 2025).
The playbook runs on moments like these, every one from your clinician-approved script library.
Day 1 · expectations4:12 PM
Week 8 · step-up eve8:30 PM
Month 7 · the plateau9:15 AMYour whole book on one board, risk-scored from silence, side effects, and refill gaps. Gigi works the wobbly ones; your staff only step in when it matters.
Dana K inbound: "been vomiting since last night, can't keep water down"
Red-flag rule matched: severe GI. Escalation triggered, agent stands down.
On-call paged with chart + full transcript attached.
Dr. Patel acknowledged. Patient notified her care team is on it.
Follow-up booked by staff. Note written back to the record.
Every play runs from your clinician-approved scripts, escalates to your team when it matters, and writes itself to the log.
Coaching lands before each dose step-up, the danger window. And the 11pm "should I stop?" gets a response in seconds, straight from your clinician-approved scripts: what's normal, what helps, what gets escalated.
The stall gets named as maintenance, not failure. Progress is reframed in percent lost, and your clinician gets a dose-review booking instead of a cancellation.
Reminders land before run-out and reorder requests route to your team. Savings programs get surfaced, and a step-down consult with your clinician beats a hard quit.
Missed check-ins and unopened texts trigger graded re-engagement: a nudge, then a personal check-in, then clinician outreach, before the patient is gone for good.
Underneath the plays: weekly weight logging by text (each 1% of body weight lost is linked to ~3% lower quit risk), onboarding expectation-setting, titration-day reminders, and win-back contact for patients who already left.
Gigi runs the predictable, your team handles the clinical. It never adjusts a dose, never gives medical advice, and adds nothing to your staff's to-do list until something genuinely needs them.
Gigi says only what your clinical team has signed off on. No improvised medical advice, ever.
Pancreatitis symptoms, severe vomiting, hypoglycemia signs: routed to your staff immediately, never handled by the agent.
Consent, STOP handling, and quiet hours enforced by the engine itself, not by a policy document.
Timestamped, auditable, attached to the record. You can always show exactly what was said, and when.
Month 12, still on therapy.
✓ The outcome every play above is built forStart patients from the book you already own. Keep them through month twelve. Win back the ones who slip. Priced so the math does the selling.
Gross profit = program fee minus drug cost. US cash-pay programs typically charge $250-500 a month, landing $230-400 of gross profit. Canadian and lean telehealth-style programs often run lower. Set the slider to your own margin.
Every $1 you put into Gigi comes back as $5.38 of kept margin.
Why two in three? Because our AI already converts at that rate for existing clients. The math above assumes Gigi matches it on your book, keeping two of every three would-be quitters on therapy.
Med spas already pay $20-30 per patient per month for generic engagement tools. One saved patient covers 20-25 patients of Gigi, and every quit you prevent also saves the $200-500 it costs to replace them.
Gigi enrolls your patients in an afternoon. $12 a month each, no setup fee, cancel monthly.