Camellia
A kombucha lab runs on precision: every jar, every reading, every recipe. We built the operating system around her process, not the other way around.
Fermentation forgives nothing.
Kombucha is alive. Every batch moves through stages, from day zero to first fermentation, harvest, cold crash, flavouring and carbonation, and each stage has its own readings, its own limits and its own timing. Miss a reading, skip a baseline or lose track of which starter went into which jar, and a whole batch can be lost.
On top of the lab sit the recipes, the cleaning and tasting logs, the stock of bottles and ingredients, and the customers waiting on the next delivery. A notebook and a spreadsheet can hold some of it. None of it talks to the rest.
Her rules, her recipes, her jars.
Our first pass modelled a factory. Big tanks, dozens of tables, a process that looked right on paper. Her lab is glass jars on shelves, and her process is her own. So we started over, by asking.
Golden Sands OS is built on her standard operating procedures, the ones she uses across every recipe and every batch. Each stage carries her limits, each reading is checked against them the moment it is entered, and the system adapted to the way she already works.
It adapted to her. She did not have to adapt to it.
About two weeks for the rebuild.
- August 2026A first version. Built ahead of the business, so we stepped back and listened.
- Early SeptemberRebuilt with Camellia around her real lab: jars, stages and her own limits. Live within days.
- Mid SeptemberCleaning, tasting and temperature logs, then stock that updates itself as she produces.
- Late SeptemberOrders and customers in the same system, and a brand system for the business.
A lab that knows itself.
The lab drawn the way it really is, each jar with its stage and its last reading.
Each measurement compared with her own limits for that stage, instantly.
The steps she follows for every recipe, so the process is the same every time.
Walks her through measurements, tracking and the day, short and in her language.
Ingredients and bottles update as batches move.
Who ordered what and when, next to the lab that makes it.
The process lives in the system, not in memory.
For anyone who makes things in batches.
What we learned in this lab carries to every small producer with recipes and batches: breweries, bakeries, chocolate and coffee roasters, cosmetics and skincare makers. Interview the maker before modelling anything, lock the facts into the data, keep the agent short and practical, and let the system follow the recipe, not the other way around.
Bring the business as it is.
We'll tell you, honestly, what we would build first and what we wouldn't. No deck, no pitch.
Start in 30 seconds