Methodology & Science
The Science Behind
AI Wellness Intelligence
MyCalAgent's AI is grounded in established research across nutritional science, behavioral health, cognitive performance, and machine learning. Here's how our methodology works — and where its limits are.
AI Meal Photo Analysis
Computer vision models identify foods from photos by recognizing visual patterns in color, texture, shape, and context. Portion estimation uses depth inference and reference objects within the frame. Nutritional values are derived from recognized ingredients mapped to established food composition databases.
⚠️ Accuracy varies based on image quality, food presentation, and cuisine type. Estimates are approximations, not precise measurements.
Wellness Pattern Recognition
Pattern recognition in wellness data uses time-series analysis to detect recurring signals across meal logs, hydration entries, fasting windows, and habit tracking. Statistical correlations identify which inputs associate with observable output patterns — such as energy, focus, or mood — over weeks of consistent logging.
⚠️ Correlation does not imply causation. Patterns are probabilistic insights, not clinical diagnoses.
Food & Mood Research
The relationship between diet and mental/physical state is supported by research in nutritional psychiatry and cognitive neuroscience. Key mechanisms include: blood glucose fluctuations affecting energy and mood, gut-brain axis signaling, neurotransmitter precursor availability from dietary amino acids, and inflammatory responses to certain dietary patterns.
⚠️ Individual responses to food vary significantly. MyCalAgent's insights are personalized to your data, not clinical prescriptions.
Hydration Science
Research consistently shows that mild dehydration (1–2% of body weight) impairs cognitive performance, reduces short-term memory, increases perceived effort, and elevates fatigue. MyCalAgent tracks daily fluid intake and surfaces patterns between hydration levels and user-reported or inferred energy signals.
⚠️ Optimal hydration needs vary by individual, climate, activity level, and health status.
Intermittent Fasting Patterns
Research on intermittent fasting (IF) indicates potential benefits for metabolic health, insulin sensitivity, and cognitive clarity for some individuals. Common IF protocols — 16:8, 5:2, and time-restricted eating — show varying outcomes across different populations. MyCalAgent tracks fasting windows and correlates them with logged energy and nutrition patterns.
⚠️ IF is not appropriate for all individuals. Consult a healthcare professional before starting a fasting protocol.
Data Privacy in Our Research
MyCalAgent is privacy-first by design. We do not sell user data or use personal wellness data to train public AI models without explicit consent.
Any internal model improvement is done with anonymized, aggregated signals — never with identifiable personal data. Users retain full control over their data and can request deletion at any time.