Engage / Lifecycle
Muscle-preservation protein and resistance coaching loop
Losing weight on the medication is the easy part — the agent makes sure it's fat coming off, not the muscle the patient needs to keep it off.
What it does
On a GLP-1, appetite drops so far that a meaningful share of the weight lost can be lean muscle, not fat — which slows metabolism, weakens the patient, and sets up the regain that undoes the whole program. Most patients never hear this until it has already happened. The agent runs a muscle-defense loop tied to where the patient is in treatment: it tracks a daily protein goal, estimates protein from meals the patient describes in plain language, nudges resistance training, and watches for signs that strength or function is slipping. It is nutrition and movement coaching, not clinical care — anything that looks like a clinical problem goes to the care team.
How it works
- 1Trigger. the medication stage advances or a behavioral signal fires — protein logged short of goal for several days, a missed resistance session, or a free-text meal the agent can score for protein.
- 2Decision. the agent weighs the signal against the patient's memory — protein target, recent intake, training pattern, treatment stage — and picks the next-best move; a judge gates each unprompted message for tone and safety.
- 3Action. a coaching touch lands on push, chat, WhatsApp, or in-app — a protein top-up suggestion from foods the patient already eats, a short resistance nudge, or quick reinforcement when they hit the goal.
- 4Follow-through. intake and training are written back to memory so the loop learns the patient's pattern; if reported strength or function is not improving, the agent flags it to a clinician with context; impact is measured against a holdout so the program sees the retention the loop earns.
Configuration
How the agent is wired for this use case.
- App backend · read treatment stage, protein goal, logged intake and training history
- Nutrition knowledge base · estimate protein from a free-text meal and surface protein-dense options the patient already eats
- Patient memory store · read intake and training patterns; write new logs and wins back for the next touch
- Messaging channel · send the judged protein, resistance, and reinforcement touches on push, chat, WhatsApp, or in-app
- Clinician escalation queue · flag stalled strength or function to a clinician with context
What you need
The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.
Signals
treatment-stage events, daily protein logs and goal, free-text meal entries, resistance-session logs
Data
protein target and treatment stage, intake and training memory, messaging consent
Guardrails
judge gating on every unprompted send; nutrition-coaching-only scope, never dose or medical advice; strength or function concerns escalate to a clinician; frequency caps; suppression once the patient is on track; holdout measurement of retained months
Metrics it moves
- adherenceup, as patients who feel stronger and see the program protecting their results stay on protocol
- churndown, because preserved muscle protects against the regain that drives patients to quit
- dau-mauup, through a daily protein-and-training loop that keeps the patient in the app
- ltvup, with each retained month compounding on already-spent acquisition cost
Related use cases
GLP-1 muscle-preservation supplement upsell
the supplement-attach motion, where this card is the behavioral coaching loop behind the same risk
Lean-mass plateau window save
the later window where muscle-loss worry meets a true plateau
Non-scale-win progress logging and celebration
the motivation loop that logs the strength and energy wins this coaching creates
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