From a photograph of your menu to a priced catalogue
A vision agent reads an existing menu into structured items. A copywriter agent drafts descriptions from the real recipe. A nutrition agent estimates the macros per serving.
SEE IT IN ACTION
Everything generated lands as a draft on the edit form. Nothing is published until a person saves it.
OVERVIEW AND BENEFITS
The tedious half of setting up a menu
Typing a hundred dishes into a new system is the single biggest reason a menu never gets finished. These three agents remove the typing without removing the review.
Import from a photo
Upload a picture of your printed menu and get back structured items with categories and prices to review.
Descriptions from the real recipe
The copy is generated from the dish’s own ingredients, category and dietary flags — not from its name alone.
Nutrition without a lab
Per-serving macros estimated from the recipe’s ingredients and quantities.
WHY IT MATTERS
A menu nobody finished entering is a menu nobody can sell from
Descriptions, allergens and nutrition are the fields that get skipped under time pressure — and they are exactly the fields the digital menu needs to be useful to a guest.
3menu agents — read, write, estimate
1photo to a reviewable catalogue
0fields published without review
PRODUCT DETAILS
Three agents, three buttons
Each sits on a screen you already use, and each hands back something you edit before it is saved.
Menu photo analysis
A vision agent extracts a structured catalogue from an uploaded menu image.
Description generation
The Generate button on the description field, fed the item’s recipe ingredients, category and dietary flags.
Nutrition estimation
Per-serving macros from the recipe quantities and each ingredient’s own nutritional data.
Always a draft
Generated text and figures are sanitised and placed in the form for you to accept or rewrite.
Generated content sits on the item, so variants and combos inherit it as they already do.
Generated from the recipe, not from the name
A description written from a dish title is guesswork. These agents are handed the actual ingredient list, so what comes back describes the dish you serve.
Ingredients in, copy out:the prompt is built from the item’s own recipe rather than its label.
Nutrition from quantities:each ingredient’s nutritional data and its recipe quantity drive the estimate.
Sanitised before it lands:the action trims and validates whatever comes back before it reaches the form.
COMPARISON
Vestibel vs. typing it all in
Capability
Vestibel
By hand
Generic AI tool
Menu photo to structured items
Descriptions built from the recipe
Nutrition from ingredient data
Lands in the catalogue you sell from
Reviewed before publishing
CUSTOMER STORIES
“
Vestibel gave us one guest record across rooms, restaurant and spa — check-in time is down 40% and we stopped losing restaurant charges entirely.
Dinesh Weerakoon · General Manager, The Banyan House
Placeholder story — no customer-story content exists yet.
A 30-minute walkthrough with your rooms, rates and outlets — no slides.
FAQ
FAQ
No. Every generated value arrives as a draft in the edit form. It becomes part of your menu only when you save it.
It is an estimate derived from your recipe’s ingredients, their quantities and each ingredient’s own nutritional data plus a serving-size hint. Treat it as a starting point to review, not a laboratory analysis.
You review the extracted catalogue before importing it, so anything misread is corrected at that step rather than landing in your live menu.
No — that is the point. Enter the recipe and generate the description from it, then edit the wording to your own voice.