The Courage to Know
Honor what you've built, without carrying it all yourself.
Look. Learn. Respond. Evolve.aunory turns your data into the next right action, so running your business takes less out of you, not more.
No account or card needed to see your first finding.Or create an account — 30 days of the full plan, free →FOOD COST$2,150 /mo
Chicken went up 22% in July. Three dishes are now losing you money.Do this: ask Sysco to match the June price. Email drafted.Review emailHow it works
From your data to your next move, in four steps.
Connect once. aunory does the data work and keeps watch, so you don't have to.
- 01ConnectLink the system you already use, or drop an export from any platform that lets you download your data.
- 02Learnaunory cleans and checks your records, links customers, sales and invoices across your systems, then learns what normal looks like for your business.
- 03PredictIt spots what's about to cost you money before it shows up in your bank balance.
- 04ActYou get the fix in plain language, with the message already drafted.
Pick a business like yours
See what aunory finds in real data from your industry.
CRITICALRestaurant · food cost
An example from a business like yours$2,150 a month$25,800 a year if nothing changes
Chicken went up 22%. Three dishes still have March prices.
Marsala, Piccata and Parm sell 1,430 plates a month. Their plate cost rose from 28% to 34%.
aunory recommends asking Sysco to match the June price. The email is already drafted.
AT RISKFriday voids 3× the other location$1,240
MONITORTuesday lunch down 18%$1,120
What aunory did in the background
- 01LookRead 12 months of Toast sales and Sysco invoices, and linked every dish to what goes into it.
- 02LearnSaw chicken jump from $2.84 to $3.47 a pound in July while three menu prices stayed the same.
- 03RespondWorked out the cost, chose the simplest fix and drafted the email to Sysco. Autopilot can send it, and always asks first.
- 04EvolveKeeps watching. Your next briefing shows whether the price came back and what you kept.
See what aunory finds in your dataCreate an accountConnect your system or drop an export. Your first finding in minutes, no account needed.
After the first answer
aunory keeps watch, so you don't have to.
Your briefing
What changed, what it costs and what to do first. Daily, weekly, or only when something is critical: you choose. By email and in the app.Ask aunory
Ask in plain words: "Why was Tuesday slow?" The answer comes from your own records, with the numbers behind it.AutopilotCOURAGE
Sends the drafted emails and texts for you, only with your permission. Every action is logged, and you can switch it off any time.How the answers are made
Answers you can check, explained in plain English.
Signals come from checks run on your actual records. Forecasts come from statistical models. The AI writes the explanation. It never makes up the figures.
- 01Your dataYour real sales, invoices and visits. Cleaned, de-duplicated and checked before anything is calculated.Toast checks18,420Sysco invoices33History12 months
- 02Statistical modelEstablished methods work out what's driving your numbers and where they're heading.Regressionthe trend lineSeasonal and trend patternsseasonal forecastYour own track recordlikely range
- 03Confidence rangeEvery forecast shows how sure we are, not one number presented as fact.Food cost next month: 31.9%
Likely between 31.2% and 32.7% - 04Plain-English recommendationWhat to do, what it's likely worth, and the message already drafted.aunory recommends raising Chicken Marsala by $1.50.Likely worth $690–$930 a month.
| The method | What it does for you | Where you see it |
|---|---|---|
| Linear regressionTrend fitting | Fits a straight line to your own history to tell a real upward or downward trend from normal month-to-month bounce. | Tested for every forecast |
| Seasonal and trend patternsPicked only when they beat a simple guess | Checks whether your slow months and busy months repeat on a yearly cycle — the slow January, the December rush — strongly enough to beat a simpler forecast. If it does, that pattern shapes the forecast; if it doesn't, aunory doesn't force it. Seasonal forecasts need about 24 months of history; with less, aunory tells you and uses a simpler method. | Predictions and next-month outlooks |
| Your own track recordConfidence ranges | The likely range comes from how far off this same method has actually been on your own past months — not a guess or a generic percentage. | Every forecast and every "likely worth" figure |
Signals never fire without real dataA signal only appears when the records that prove it are in your data.
Forecasts show a rangeWith six months of history or more, every prediction comes with a likely low and high. Less certain means a wider range.
Thin data, no guessWhen there isn't enough history to be sure, we tell you what's missing.
Start with one
clear answer.
Connect your system or drop an export. No card needed.