NOTES FROM THE BUILD

What we are learning, written down as we go.

Food data, diet modelling, progression engines, and the decisions that looked obvious until somebody ate something.

A lifter setting up over a loaded barbell on a gym floor

Product28 August 20267 min read

The gap between the bar and the bowl

Your training app does not know what you ate. Your food app does not know what you lifted. Every interesting question lives in the space between them.

Read it

Ingredients laid out on a wooden cutting board: avocado, tomatoes, eggs, mushrooms, spinach, rocket and spring onions

Nutrition2 September 20266 min read

Nobody weighs a roti

Food apps ask for grams. People eat in rotis, katoris, slices and cups. We built the catalog around the second one, and it changed the schema.

Slices of cucumber, apple, lemon, celery, kale and spinach arranged in a grid on a pale green background

Engineering8 September 20268 min read

Diet is not a ladder

Vegan is not a stricter vegetarian, and Jain is not a stricter vegan. Model diet as one slider and you will quietly serve people food they will not eat.

A lifter under a loaded barbell, photographed in black and white

Training11 September 20267 min read

Progression without a spreadsheet

Every serious lifting programme is a set of rules that somebody maintains by hand. We wrote the rules down and let them run.

One place for the bar and the bowl.

Built in the open. Leave an email and you get one message the day it ships.