Reports
Numbers that answer
when you ask.
Sales, labor, purchasing, and food-usage reports per location, built from numbers that combine every night. A built-in assistant has read all of them, so you can ask in plain language and get an answer that cites the receipt.
$3,184 across 9 deliveries, down 6% from the prior two weeks. Tomatoes drove most of the change.
Also available from Claude, with owner-approved actions.
Four report families, zero dashboard sprawl
Sales
What sold, when, and what it brought in.
Labor
Hours and cost against the sales they served.
Purchasing
Every dollar out the door, by vendor and week.
Food usage
What the kitchen actually consumed, priced.
Data collects itself
Sales, hours, and purchases flow in from what you already run and combine overnight. No exports, no spreadsheets.
Reports stay per-location
Each store reads its own numbers, permissioned by role. Owners compare across all of them.
Ask anything
The assistant answers from the same data and cites where each number came from. Connect Claude or ChatGPT through Integrations to ask from anywhere.
Four report families
Sales, labor, purchasing, and food usage, per location. Each one answers an operator question: what sold, what did labor run, what did we buy, what did we use.
A weekly report that reads like a manager wrote it
The flagged items come first: the price that moved, the variance that widened, the labor that crept. What needs your attention, before the numbers you already expected.
Built-in AI chat
An assistant grounded in your own data. Ask why food cost rose two points and the answer cites the invoices that moved it.
Owner-approved actions
The assistant can prepare changes, but nothing executes without your explicit confirmation. Read access is the default, and every proposed change shows exactly what will happen first.
Reports built for the pass, not the boardroom
Each report answers an operator question: what sold, what did labor run, what did we buy, what did we use. Sales, labor, and purchases combine overnight, so the number you read in the morning is the real one.
- Per-location, with role-based access
- Sales, labor, and purchases combined nightly
- Theoretical vs. actual food usage, variance flagged
- Export anything
An assistant with receipts
Because the assistant reads the same priced invoices and reports you do, its answers cite real numbers, and its proposed actions wait for your confirmation. Ask what changed this week and it points at the lines that changed.
- Grounded in your invoices, sales, and counts
- Answers point at the exact lines that moved
- Prepare-and-confirm flow for any change
- Want it in Claude or ChatGPT? See Integrations
4
report families, no dashboard sprawl
1
question replaces an hour of spreadsheets
0
actions taken without your confirmation
Good to know
What can the assistant actually do?
It answers questions grounded in your own invoices, sales, and counts, and can prepare changes that wait for your explicit confirmation before anything executes.
Is my data used to train models?
No. Your prices, invoices, and counts stay yours, and never train anyone else's models.
Who can see which reports?
Access is role-based per report family, so bookkeepers, chefs, and managers each see exactly what they should.
Can my accountant get the reports?
Yes. Every report exports, and bookkeepers can have role-based access of their own. They see the numbers without touching ordering.
The full suite
It all runs on the same connections.
Get started
Sixty days free.
Setup done for you.
We're onboarding a limited number of early restaurants. Tell us about yours and we'll reach out within a day.
- 60-day free trial, no card required
- Vendor connections set up with you
- Cancel anytime; your data stays yours