brewOS

Production software
for people who brew.

Recipes, brew day, the cellar, packaging, sensory analysis, and when each beer has to be brewed. In five languages, on the phone in your pocket.

A brewmaster beside a fermenter

It's Thursday. The sheet taped to FV3 says day nine in marker, and nobody is sure if that's nine days since brew day or nine days since it was moved. The whiteboard by the door was right on Monday. Your head brewer is off today, and his readings are in his notebook.

That's what happens when a brewery grows past what one person can keep in their head.

A brewer writing in marker on a fermentation sheet in a plastic sleeve, cable-tied to a fermenter

“Somebody else is taking the readings now.”

Six in the morning, one person, every tank. Gravity, temperature, pressure, a look at the krausen. The numbers that tell you what every beer in the building is doing — and they end up in a notebook in somebody's back pocket.

Readings go in from a phone at the tank — gravity, temperature, pH if you want it — and each one carries the name of whoever took it, from their login.

Someone enters 1.012 where they meant 1.021. Both believable, which is why nobody catches it, and why the beer nearly gets crashed three days early. You fix it, and the first number stays on the record beside the correction. Nothing you write down quietly goes away.

No signal back between the tanks? The reading waits on the phone and arrives when it reconnects.

A brewer at a fermenter, taking a reading on her phone

“Nobody's looked at FV4 since Friday.”

A lot of bad beer starts here. Not with a disaster — with a tank nobody opened the log on, a temperature that drifted over a weekend, a ferment that stalled three days ago.

BrewOS opens on what needs you. A tank with no reading for two days in primary. A temperature more than 2 °C off what the recipe asks for that stage. A ferment that's dropped less than 0.2 °P over two days and three readings while it's still well short of its final gravity — and it says for how many days.

It tells you which tank. It doesn't tell you what to do about it — you know things it doesn't. And where your recipe states no target, it raises nothing rather than inventing one.

A brewer with her palm flat against the side of a fermenter, looking up the length of the tank

“This batch doesn't taste like the last one.”

Or it does, and you can't prove it.

Most breweries don't run a panel. They run the brewmaster, who tastes it and decides — which works until the morning somebody disagrees and the room settles it by who spoke first.

Run a triangle test and let the numbers answer. Tasters join from their own phones with a code and no login, and every sample is blind-coded. Seven methods, from a straight difference test to the DLG scheme, each worked out from the answers.

What goes in front of the panel is up to you — this batch against the last, a trial against the house version, your beer against the one everybody keeps comparing it to.

And if you've never run one properly, the guide is right there — method by method, with worked examples, reference tables and coaching notes. In five languages.

Four tasters at a table in a clean room with coded plastic cups and answer sheets

“A keg came back and we don't know which tank it came from.”

Type the lot number off the label, exactly as it's printed. You're looking at the tank it came from, the day it was packaged, and where the rest of that lot went.

Stock in the cold room isn't a number somebody keeps overwriting — every keg on the shelf traces back to the run that filled it and everything that's happened to it since, with its packaging date and its age in days.

A brewer rolling a returned keg across the brewery floor

“We're arguing about whether to buy another fermenter.”

And nobody has a number to settle it with, so it gets decided in November, under pressure, on a feeling.

Put in the orders and the dates you promised them. BrewOS works backwards through each recipe's stages — against your brewhouse hours, every tank the beer passes through on its way, your packaging line, and the stage times measured on your own batches — and tells you where you run out.

Not capacity exceeded. You are not short of brewhouse. You are short of tank time.

By the time an order feels urgent, the beer needed to be in a tank weeks ago — three or four weeks for a straightforward ale, two to three months for a lager given its time. This is that conversation, a year before you're forced to have it.

Read the long version in the journal

Three brewers in aprons looking up at a fermenter, one pointing

“The whiteboard is wrong again.”

It was right on Monday.

The plan publishes onto a schedule your crew actually reads — week, month, quarter, filtered by person; on a phone it opens on just their own tasks. Move a brew on the floor and the plan knows it moved, who moved it, and when.

There's a wall board too, for the screen by the door — and in the morning meeting your crew moves a task to Now, Next or Done with a PIN. It's what the whiteboard has been trying to be.

A brewery crew around a table while two of them work through the week on a whiteboard

Recognise your own brewery in any of those? We're choosing twelve to run BrewOS with us from the start — what that involves. Applications close 30 November.


One thing we ask for

The recipe.

Not just the grain bill — the mash steps, the lauter, the boil, the knockout, the fermentation. The targets you'd actually judge the beer against.

Fill it in once, and brew day shows you're under on pre-boil gravity while there's still something you can do about it — not the next morning, when the beer is in the tank. The cellar tells you the beer is behind where the recipe had it. The planner works backwards through stages it knows the real length of.

It's the one place BrewOS has an opinion.

A taster answering on her own phone beside two coded cups
Something we're designing

An SOP companion.

Every brewery knows it should have SOPs. Most have three, all out of date.

Standard Operating Procedures without the blank page — a starting outline, the published references beside it, and a one-pager for the wall by the tank.

Being designed. Not built, and not a date.


Why it exists

I didn't set out to write software. I set out to stop running the brewery from memory.

The things that went wrong were never dramatic. A reading that stayed in somebody's notebook instead of going on the batch. A tank I'd counted as free. Two people who each thought the other was doing Tuesday's brew. A beer I was sure had changed, with nothing either way to prove it.

Underneath all of them was the same thing. The brewery ran on what was in my head, and what's in your head doesn't survive a holiday, a new hire or a bad Thursday.

So I started writing it down properly. Not a grain bill and a hop schedule — the process. The rests, the lauter, the knockout, the targets you'd actually judge the beer against. Then who does what and when. Then what the plan does when reality moves, because it always does.

None of it is new. Every rule in BrewOS is something good brewmasters already do. What they have that most small breweries don't is the discipline written down and within reach — so the practice belongs to the brewery rather than to whoever happens to be holding it.

That's the whole idea. Stop doing it that way because it has always worked. Do it that way because you know why it works.


Come in. The kettle's on until the end of November.

We're choosing twelve breweries to run BrewOS alongside us. The price you start at is the price you keep, the setup is done with you rather than handed to you, and you get the brewmaster's direct line instead of a ticket queue.

Applications close 30 November