The Karu Inbox: One Intake Flow for Invoices, Recipes, Menus, and Sales
How a single document inbox turns PDFs, photos, spreadsheets, and pasted text into structured drafts ready for review.
Margin systems fail when every document type needs its own upload ritual. Karu's Inbox is the intake center for the files a food business already has: supplier invoices, recipe sheets, menus, price lists, sales exports, photos, PDFs, and pasted text.
One front door for messy inputs
Owners should not need to guess whether a file is an invoice job, a recipe job, or a menu job before uploading. The inbox accepts the mess and routes it into the right draft type.
That keeps the first session lightweight: drop files, review drafts, approve truth.
Parsing is the start, not the finish
After upload, Karu extracts structured candidates: supplier lines, ingredients, products, units, prices, and confidence scores.
Those candidates land in the Review Center. Official recipes, supplier prices, and product links only update after approval.
The inbox connects to margin action
Once approved data exists, the same intake history explains why a margin moved: new invoice, recipe change, sales upload, or packaging update.
That is how Karu turns document intake into decisions — fix, push, or reprice — instead of another archive folder.
Operator checklist
Upload invoices, recipes, menus, and sales through the same intake flow.
Review extracted drafts before they affect official costs.
Keep source files tied to approved price and recipe changes.
Use approved data to rank margin actions, not just store files.
Sources
Related guides
Operations
What to Upload First: A 3-Document Margin Diagnostic
You do not need a perfect data room to find the first margin leak. Three real documents are usually enough to see what is confirmed, what is weak, and what to review next.
AI
Invoice Extraction with AI: Turning Messy Supplier Documents into Usable Costs
How multimodal AI can reduce the pain of invoice data entry while keeping humans in control.
AI
AI Data Cleanup for Restaurants: From Messy Inputs to Reviewable Drafts
How AI can organize recipes, invoices, menus, photos, and text without removing human approval.