your food tracker, now with carbon efficiency

>200 million people track their macros. fewer than a million track the carbon cost of eating. no tool does both.

v0 prototype · emissions model v0.2 · looking for collaborators

see how it works → get involved →

diet and carbon tracking tools don't talk to each other

>300M^{1 2}

users worldwide

I am changing my diet and going to track my nutrition.

I wonder how that will affect the environment.

Have you tried a carbon calculator?

Yeah but if I'm already tracking my meals…

If only MyFitnessPal had built-in emissions info like Google Flights 🤔

october 2025. we started building the next monday.

emissions data nudges demand

google flights put a CO₂e figure next to every fare. studies since have measured a 4–12% shift toward lower-emissions options^{1 2}. nobody has run that experiment on food.

4–12%

measured shift, air travel

we brought meal tracking and carbon accounting together

System architecture: the Carbon Foodprint mobile app talks to cloud compute, which reads and writes a Supabase database; MyFitnessPal data flows through the Terra API into Supabase, and public data sources feed the emissions model, which also writes into Supabase. — v0 architecture — tracker data in via terra, emissions model alongside it

Carbon Foodprint app, Today screen: macronutrient rings for protein, carbs and fats above a carbon impact card showing total emissions, a 7-day average and nutrition efficiency, with a warning that two meals need ingredient data.

Carbon Foodprint app, My Foodprint screen: weekly total foodprint of 15.7 kg CO2e and foodprint efficiency of 0.8 g CO2e per calorie, each with a sparkline, above a meal breakdown list.

foodprint: g CO₂eq
efficiency: g protein / foodprint
trend: ± % avg efficiency
cost-effectiveness: efficiency / cost

we deliberately don't set carbon targets. we won't ask anyone to trade nutrition for climate.

the model is the hard part

Sankey diagram for a sample meal: 40 g protein, 40 g fat and 45 g carbohydrates flow into inferred ingredients — 150 g chicken, 50 g egg, 200 g rice, 30 g sauces and 30 g cucumber — which combine into a total of 1.85 kg CO2eq. — sample: hainanese chicken rice with braised egg. macros in, ingredients inferred, emissions out.

our v0.2 model maps ingredients to g CO₂eq using Poore & Nemecek (2018) — 38,700 farms, 119 countries, farm-to-retail — extended with Ecoinvent and a hand-built set of common ingredients the public datasets miss.

known limitations

— public data is thin, aggregated, and years old
— coverage stops at the major food categories
— huge real variability is hidden inside single averages
— processing and logistics stages are missing or inconsistent (drying, roasting, fresh vs frozen)
— system boundaries and methodologies don't agree across sources
— meal data arrives as macros, not ingredients, so we infer

we publish the mapping and the uncertainty rather than hide them. if you work with primary LCA inventories, product-level footprints, or supplier data — including data that can't be published — we'd like to compare notes.

methods note (in progress) → alex-wzm@outlook.com capstone deck, pdf ↓

where you fit in

validate the problem

is there real demand for a meal-level foodprint?

improve the model

better coverage, honest uncertainty, regional factors

meet people where they eat

delivery apps, cafeterias, nutrition platforms

we're looking for

LCA & food supply chainnative app dev (RN or Flutter)UX designdietitiandata partners

shipping log

2025-12-04: capstone presentation
2025-11-25: emissions model v0.1, sankey prototype
2025-11-12: app screens designed
2025-10-16: first pitch, climatebase cohort 5

team

Alex Wong Software Engineer

Kira Aglio MSc Environmental Tech

Di Xu LCA Consultant

Andrew Montross Operations Manager

alex-wzm@outlook.com