Outlier resistance
One unusual weigh-in should not create a large estimate or target change.
Adaptive TDEE
Total daily energy expenditure (TDEE) is an estimate of the energy your body uses across a day—the “calories out” side of calorie balance. Quevro is being built to learn it gradually from your own longitudinal data, rather than treating a one-time calculator or wearable number as permanent truth.
The starting point
Quevro plans to calculate an initial basal metabolic rate using the Mifflin–St Jeor equation, then apply a general activity factor. This provides a reasonable onboarding estimate while the app has no personal history.
It is not a direct measurement of metabolism. Personalization begins only after enough usable intake and weight observations accumulate.
Signal hierarchy
Quevro’s planned model prioritizes what can be observed across time. Activity helps explain the day, but intake, trend weight, and time do more of the long-term learning.
Complete daily totals provide the energy-in side of the relationship. Missing or partial logs reduce confidence.
A smoothed trend helps distinguish direction from normal water and measurement noise.
Patterns observed across 7, 14, 21, and 28-day windows are more informative than isolated days.
Weigh-in coverage, intake completeness, missing days, and data flags determine how much trust to place in the estimate.
Steps, workouts, and available energy data can explain context, but should not be double-counted or treated as exact.
The core relationship
Quevro plans to estimate calories out from a baseline, then refine that estimate using the relationship among logged intake, smoothed weight trend, and elapsed time. Calories in minus estimated calories out produces a working view of net energy balance: likely deficit, near-balance, or surplus.
This is a model, not calorimetry or a promise of weight change. Short windows can be distorted by hydration, glycogen, sodium, digestion, illness, and logging error.
Confidence model
Levels are planned to respond to observation duration, log completeness, weigh-in coverage, agreement across windows, and recent data quality.
The estimate is still anchored mostly to onboarding inputs.
Some usable observations exist, but gaps or limited time constrain the model.
A steadier intake and weight-trend pattern is becoming visible.
Multiple windows agree and recent data quality remains strong.
Built-in restraint
One unusual weigh-in should not create a large estimate or target change.
Proposed estimates are planned to stay within physiologically conservative ranges.
New information is incorporated over time instead of replacing the prior estimate abruptly.
Large calorie-target changes are intended to arrive gradually with an explanation.
Incomplete days lower confidence and should not masquerade as unusually low intake.
Estimates, inputs, confidence, and quality diagnostics are planned to be retained for context.
Limits matter
Quevro cannot directly measure metabolism or determine why weight changed. Even a high-confidence estimate can be affected by unlogged intake, fluid shifts, travel, training changes, illness, medication, pregnancy, and other factors.
Quevro is not a medical device and is not intended to diagnose, treat, cure, or prevent any disease. Calorie targets and projections are informational estimates, not medical advice. People with a medical condition, a history of disordered eating, pregnancy-related needs, or concerns about nutrition or weight should consult a qualified health professional.
Questions about the approach? Contact us