Accuracy

Reliable numbers with visible assumptions.

Each entry is grounded in dependable food data and presented in a way you can validate quickly.

Method

USDA source

Food matching uses USDA FoodData Central as the primary macro data source.

Transparent matching

You can inspect which item was selected and why that match was used.

Portion assumptions

Approximate descriptions are converted into a clear serving assumption in the log output.

Decision-ready totals

Calories and macros update immediately so you can adjust the day in real time.

Food types we log

Survey foods

Generic foods from USDA FNDDS with standardized portions and reliable nutrition structure.

Branded foods

Packaged or commercial products from USDA branded data when generic survey matches are not the best fit.

Quick-log foods

Direct macro entries used only as a fallback for custom or complex foods when USDA matching is not sufficient.

How AI logging works

  1. Break the meal into searchable components.
  2. Search USDA foods first, prioritizing survey foods and then branded foods when needed.
  3. Select the best match and portion, then calculate macros from that portion and quantity.
  4. Use quick-log only when USDA search and reasonable generalization do not produce a good match, or when the food is truly custom/complex.

Common accuracy questions

How is a food match selected?

Ziva chooses the closest USDA-backed match from your description.

What about vague portions?

A best-fit portion assumption is included so the math stays explicit.

Is this medical-grade diagnosis?

No. It is designed for day-to-day nutrition guidance and tracking.

Ziva is a practical nutrition tracking tool and does not replace professional medical advice.