We build the measurement layer that makes Indian dietary assessment work, starting with the portion units Indian kitchens actually use.
Food composition databases store nutrient values per 100 grams. Indian households do not measure food in grams. They measure in katoris, rotis, ladles and tumblers. Almost every dietary tracking tool available in India bridges that gap by asking the user to guess, or by silently applying a Western default serving size.
The gap is not a rounding error. The same 150 ml katori holds very different weights depending on what is in it and how it was cooked.
| Household unit | What has to be established |
|---|---|
| katori | size class, then grams by food density — thin dal, thick rajma and dry sabzi differ substantially at equal volume |
| roti / chapati | diameter and thickness, and flour type: atta, maida, bajra, jowar |
| dosa | plain, masala or paper — roughly a threefold weight range |
| idli | count-based, but regional size varies by about 1.5× |
| colloquial quantity | "thoda", "aadha katori", "ek chammach" still resolve to a number |
Multiply those units across several thousand regional dishes and every cell has to be measured rather than estimated. That table does not exist as a public resource. Building and validating it is our current work.
A validated mapping from household serving vessels to weights, resolved by food class and cooking method, across Indian regional cuisines.
People describe a meal in their own language and in their own units. Speech and language handling for Indic languages, not a translated interface.
Every user correction to an estimated portion is a data point. Aggregated, it produces empirical portion distributions across regions.
Indian food composition data is the product of decades of publicly funded scientific work. We treat it accordingly.
Our foundational references are the Indian Food Composition Tables 2017, published by ICMR–National Institute of Nutrition, Hyderabad, and the Indian Nutrient Databank, developed by Anuvaad Solutions. Neither institution is affiliated with Ariyah Labs, and neither has endorsed this work.
Ariyah Labs is at the data foundation stage. We are assembling and validating the portion ontology before building consumer product, because the ontology is the part that determines whether anything built on top of it is accurate.
We are actively seeking data partnerships and licensing conversations with nutrition research institutions. If your organisation holds Indian food composition or dietary intake data, we would value a conversation.