Every ingredient price in your city. Every menu price on your street. And the math on your dishes, done.

Keep it. Your system deducts stock from a recipe, with a cost that someone typed in when the dish was created. It does not know what pork belly cost at the market on Tuesday, and it does not re-cost anything when a price moves. We send it this week's prices for the ingredients you name, every week, as a file or a feed, so the cost it already holds becomes a cost that updates. If you would rather, our app does the costing and you export the number back.

A chatbot can write a plausible cost sheet. It does not know what shrimp cost at your supplier this week, what is in your central kitchen's cold room, or which fish sauce your group uses. Our AI Kitchen Manager runs fixed, tested routines on your real recipes, your real stock and real local prices, and shows its working, so your accountant can check it.

A delivery app shows the prices of the restaurants it lists, with the commission baked in, and it shows nothing for the places that sell from a shopfront, a cart or a Facebook page. We capture every menu on the street, listed or not, so you see the whole price band for your category, not the part one app chose to show you.

Then you already know the hard part: updating it. We feed it this week's prices, or make it automatic in our app and point it forward: which dish is drifting, which kitchen, which supplier quote is above the market, and what a bulk offer is worth across the group. You keep the spreadsheet; it just stops being a Sunday job.

Yes. That is the shape we call a capture you name: the ward around your outlets, the street you compete on, the ten stores your buyers use, or your own distributors' quote sheets, receipts and purchase orders, which we turn into the same priced, dated lines as our market photographs so they sit in one list. Market and street captures from December; distributor documents from January. Founding Kitchens' captures go first.

Not with the books themselves. Since the lump-sum tax ended in January 2026, groups need a record of cost with documents behind it, and a purchase list for buying at the wet market. We do not keep tax books, issue e-invoices or write the purchase list; your accountant and your POS do. What we give them is a current cost per dish with a photographed market price behind every ingredient, exported to the accountant's spreadsheet, which is a better starting point than a cost typed in two years ago.

For the ingredients you name: this week's price at each place you could buy them, with the place, the date and the photo on every line, as a spreadsheet or CSV, as JSON for your own system, or inside your Mintari workspace. For a street: a report with every menu item within 500 metres of the outlet, its price, and your dishes placed in the band. Prices come from Ho Chi Minh City's markets from December and its streets from January.

The files you have (Word, PDF, spreadsheets), photos of the printed cards and the notebooks, using our simple photo guide, or a person with a book-scanning rig at your central kitchen. You get back every dish and every sub-recipe, checked, searchable and costed, in the format you choose: a PDF for the kitchen wall, a spreadsheet for the accountant, a recipe-app file (JSON-LD or Paprika; we use the schema.org standard) for any app, or a workspace in our Mintari app (planning and budgeting for kitchens).

The language your kitchen writes in, printed or handwritten. We return each recipe in the original and, when you want it, with a translation beside it, so your kitchen and an English-speaking chef, or an overseas outlet, cook from the same card.

Every price carries its photo, so you can check it. Every recipe is graded by software; anything flagged is checked by a person. We report the error rate per source instead of hiding it. If a book comes back worse than we promised, we fix it or you do not pay.

Yes. A book you send belongs to you alone; we never build a shared library from anyone's recipes, and we never show one group's recipes, costs or margins to another. The menu prices we capture on the street are public, and yours appear in the city dataset like everyone else's; your costs never do. We delete your photos after delivery, and you can export everything and leave at any time.

One kitchen's take on a dish: its supplier, its brand of fish sauce, its portion, the step the head cook there insists on. A version keeps its link to the master, carries its own cost, and shows your language and English on one page, so the culinary director can see every difference between kitchens and decide which ones are the house style.

It keeps the costing sheet for you and points it forward. It costs every dish from this week's market prices, sub-recipes included, in every kitchen and on every channel. It re-costs everything when the market moves, shows that beside what your stock cost you, and tells you which dishes and which outlets are affected. It keeps every kitchen's version beside the master, each with its own cost. It costs a new dish before it goes on the menu and shows the street's price band for it. It sorts your buying list by place and checks a supplier's quote against the market. It tells the central kitchen what to prep this week from what is already in the cold room, and it costs a buffet, a catering order or the staff meal for a head count. And, coming next, it reads your sales, your stock and your expiry dates together to say how many cases of a distributor's offer pay back, and which kitchen they belong in.

Ho Chi Minh City from December, weekly, with menus street by street from January. Bangkok is the next city, and the groups that join us first help decide when, and which markets and streets go first. Until then, your own supplier prices work for costing.

Yes. Each kitchen gets its own account with its own versions and costs, and the culinary director and the owner see all of them side by side. We are building the group view with the first groups that join us, and we would like your view on what it should show first.

The data is priced on its own: a monthly subscription for the weekly feed, a one-off report (a street benchmark, say), a capture of what you name (a ward, a street, the stores you buy from, or your distributors' quote sheets and receipts), or a single answer to one question. The AI Kitchen Manager is a small monthly fee per kitchen account, covering cloud hosting for your book, the AI Kitchen Manager and recipe costing, with the data inside it. The Cookbook Converter is priced per page, pennies each, with a volume rate for a whole group. Founding Kitchens lock in lifetime terms on whatever they take. We quote on the demo call, with your numbers.

A Founding Kitchen is one of the first restaurant groups we build these tools alongside: your feedback shapes what we build next. In return you get lifetime terms, a say in what we capture next, the markets you buy from and the streets you sell on captured first, a 30-day head start on that data before anyone else sees it, and your name on the price sheets and the city board.

You pay for the weekly capture of one ward or one market, the one around your outlets or the one you plan to open in. We capture it first and on your schedule, you see the data for 30 days before it enters the city dataset, and the price sheets for that ward carry "captured with support from" and your name. It is a product, not an investment: you are buying the data early, not a share of anything.

The intro call is twenty minutes: who we are, what is launching in the next weeks, and what you are trying to do over the next year. If it is useful, we agree a free first test (one street, your buying list, or your recipe book) and a date for a demo call on your own material in December.

Export everything and go. No lock-in is the point.