Appetite AI
A free mobile health app built on the same AI and sensor technology that runs our closed-loop farm — predicting hunger-satiety cycles, tracking metabolic phases, and providing biological rescue protocols to help people eat at biologically optimal times. Built in Flutter for iOS and Android. Pending app store approval. No subscriptions, no accounts required.
The Core Feature: Hunger–Satiety Correlation Chart
The main screen plots a 24-hour hunger curve showing the biological correlation between meal timing, satiety duration, and hunger intensity. Each meal creates a visible drop in the curve — the deeper the drop, the longer the satiety window. The app predicts when hunger will return and schedules your next meal at the biologically optimal point before cravings spike. A green “7.0 Sweet Spot” line marks the threshold: below 7 means your body is still satisfied; above 7 means you are approaching the optimal eating window.
The curve is not a static prediction. It combines baseline circadian biology, real-time wearable data from Apple Watch (heart rate, activity level, step count, sleep quality), logged meals with macronutrient breakdown, and the user’s active fasting protocol to produce a continuously adjusted forecast. Three overlay layers are togglable: Live Adjusted (blue, incorporating all real-time inputs), Protocol Simulation (green dashed, showing predicted curve under the selected fasting protocol), and Standard Biology Baseline (orange, showing the unmodified circadian hunger pattern). Time window selectors (24h, 12h, 6h, 3h, 1h) zoom into specific periods for detailed analysis. This approach draws on research published in the NIH PubMed database on circadian rhythm influences on appetite regulation and meal timing.
Metabolic Phase Tracking
A live countdown ring tracks the user’s current biological phase based on elapsed time since their last meal. The four phases are grounded in established metabolic science: Digestion (0–4 hours, insulin active, glucose being absorbed), Post-Absorptive (4–12 hours, blood glucose being consumed, glycogen being tapped), Fat Oxidation (12–16 hours, glycogen depleted, fatty acid oxidation becomes primary fuel source), and Deep Fast (16+ hours, autophagy activated, cellular cleanup processes engaged). Each phase displays a real-time countdown to the next metabolic milestone, a description of what the body is doing, and the phase’s characteristic color coding. The ring animates as the phase progresses, providing an intuitive visual of metabolic state without requiring the user to understand the underlying biochemistry.
SOS Rescue Protocol
When hunger intensity reaches crisis levels, the SOS Rescue button (pulsing red, always visible) launches a 45-minute guided biological recovery system. The protocol begins with a Biological Triage assessment — a bottom sheet presenting two scenarios based on symptom severity:
Level 7 (Intense Cravings): Stomach growling, fixated on food, but blood sugar is stable. Action: 500ml electrolyte water or black coffee. Wait 20 minutes. This is a zero-calorie intervention that addresses the psychological component without breaking a fast.
Level 10 (System Crash): Dizzy, lightheaded, cold hands, shaking, cannot focus. Action: 1 hard-boiled egg + 250ml warm water. This triggers cholecystokinin (CCK) release from the duodenum, which sends satiety signals to the brain within 5–10 minutes.
Selecting Level 10 launches the Rescue Timer HUD — a 45-minute countdown progressing through three biological stages, each with real-time explanations of what the body is doing:
Stage 1: Mechanical Stretch (0–5 min) — The stomach senses the egg and water. The vagus nerve begins transmitting “stand down” signals to the hypothalamus. The panic begins to fade.
Stage 2: Hormonal Switch (5–20 min) — CCK and GLP-1 (glucagon-like peptide-1) hormones are released into the bloodstream. These are the biological brakes on hunger. The hormonal cascade physically suppresses appetite at the receptor level.
Stage 3: Blood Chemistry (20–45 min) — Amino acids from the egg reach the liver. Blood glucose stabilizes. The 10/10 starvation alarm physically shuts down as the body confirms that calories have arrived and the crisis is over.
Additional Features
Keto Gauge: A five-strip visual indicator estimating ketone body levels based on glycogen depletion rate, elapsed fasting time, and recent carbohydrate intake. Strips range from Negative (< 0.5 mmol/L) through Trace, Small, Moderate, to Large (8.0+ mmol/L). The gauge provides a non-invasive estimate without requiring blood or breath testing. Note: this is an estimation model, not a medical measurement.
Sensation Logging: Users log real-time physical sensations — Fullness, Headache, Dizziness, Acid/Heartburn, Coffee intake, Sleep quality, and Stomach Growling — via quick-tap chips. Over time, these data points calibrate the hunger prediction model to the individual’s unique physiology, improving forecast accuracy.
Protocol Library: Six pre-built intermittent fasting protocols with customizable meal windows: Time-Restricted Feeding at 16:8, 18:6, and 20:4 ratios, One Meal A Day (OMAD), Protein-Sparing Modified Fast (PSMF), and the 5:2 method (5 normal days, 2 restricted days). Each protocol adjusts the hunger curve simulation and meal scheduling accordingly.
Wearable Sync: Apple Watch integration syncs heart rate, step count, sleep duration/quality, and active calories every 15 minutes. This data feeds the hunger prediction model — elevated heart rate after exercise suppresses appetite differently than resting heart rate, and sleep quality directly impacts next-day hunger patterns.
Farm Integration: The app connects to AI Sustainable Foods’ direct-to-consumer infrastructure. Users can browse available products, place orders for pickup at the Route 115 farm stand, reserve items in the 25-locker vending machine, and subscribe to CSA shares. The farm page details the production systems behind the food.
From Farm Sensors to Health App
Appetite AI was not conceived as a separate product. It grew directly from the sensor data pipeline and biological prediction models running on the farm. The same approach that predicts when dissolved oxygen will drop in the shrimp tank — combining environmental sensor data with biological cycle models — predicts when hunger will spike in a human. The same approach that forecasts when BSF larvae will reach harvest weight predicts when a user will transition from Post-Absorptive to Fat Oxidation phase.
The technology stack is identical: time-series data collection, biological model overlays, real-time dashboard visualization, and threshold-based alerts. The farm uses Grafana. The app uses Flutter. Both consume the same type of data and present the same type of predictive overlays. Read more about the farm’s technology stack.
Status & Availability
Appetite AI is feature-complete and built in Flutter for both iOS and Android. The app is currently pending approval from the Apple App Store and Google Play Store. Once approved, it will be available as a free download with no subscriptions, no in-app purchases, and no account creation required. Google AdWords campaigns are approved but not yet launched pending store availability.
The interactive demo above is a fully functional simulation of the app’s core features, built from the actual Dart/Flutter source code. It demonstrates the hunger-satiety chart, metabolic ring, keto gauge, sensation logging, SOS rescue protocol with biological triage, and the three-stage recovery timer. All interactions are live — try tapping the SOS button, switching tabs, or toggling the chart legend.
All Donations Are
Tax-Deductible
509(a)(2) public charity — EIN 93-1895646 — Candid Gold Seal 2026.
Exclusive Donor and Member Benefits
— Free workshop passes
— Behind-the-scenes tours
— Priority CSA subscriptions
— Naming recognition
— Members-only updates
Frequently Asked Questions
Is Appetite AI free?
Yes. Free to download, no subscriptions, no in-app purchases, no account required. The app is funded as part of AI Sustainable Foods’ nonprofit mission to improve community food access and health literacy.
When will the app be available?
The app is feature-complete and pending Apple App Store and Google Play Store approval. The interactive demo above provides full access to all core features in the meantime.
Does the app require an Apple Watch?
No. The Apple Watch integration enhances predictions with heart rate, activity, and sleep data, but the app works without it using manual meal logging and the base biological model.
Is the hunger prediction medically validated?
The prediction model is based on published research on circadian appetite regulation, CCK/GLP-1 hormone responses, and metabolic phase transitions. However, Appetite AI is a wellness tool, not a medical device. It does not diagnose, treat, or prevent any condition. Users with eating disorders, diabetes, or other medical conditions should consult their healthcare provider before using fasting protocols.
What is the connection to the farm?
The app uses the same sensor data pipeline and biological prediction approach as the farm’s IoT monitoring system. It also integrates with the farm’s direct-to-consumer sales — users can order food from the Route 115 farm stand and 25-locker vending machine through the app.
How does the SOS Rescue Protocol work?
When hunger reaches crisis levels, the SOS button launches a biological triage (Level 7 or Level 10) followed by a 45-minute guided timer explaining vagus nerve signaling, CCK/GLP-1 hormone release, and blood chemistry stabilization in real time. Try it in the demo above — tap the red SOS button.