Customer service
Answer menu and dietary questions instantly
AI replies about allergens, vegan options and hours, then hands the guest to you to book.
Guests deciding where to eat ask the same questions again and again — is there a vegan option, are you gluten-free friendly, are you open tonight — and a slow reply sends them to the next restaurant. This workflow lets an AI agent answer accurately from the menu, allergens and hours you have described, then point the guest to you to book, so a question becomes a guest at the door instead of a lost table.
Last reviewed
Can AI answer allergen questions about my menu?
It answers from the allergen notes you have written, not from the kitchen. The agent can say which dishes suit which diets based on what you have described, and sends anything it cannot confirm to your team rather than guessing.
- Answers
- Your menu, allergen notes and hours
- Hands over
- Anything the notes do not cover
- Not yet
- Speaking for the kitchen tonight
Your customers are already asking AI
These are the kinds of prompts people now type into an AI assistant — the demand this workflow is built to serve.
“find a restaurant with vegan options near me”
“which restaurants near me are gluten-free friendly”
“is there a place nearby with a kids menu”
“find a restaurant that caters to nut allergies”
So how does a customer's AI find this at all?
These questions only turn into covers if guests find you first. Make your menu and this dietary assistant available to AI, so when a diner asks their assistant for a place with the options they need, AI can point them to you and answer on the spot.
You describe your menu, your allergen notes and your hours — and Fugentic publishes that as an MCP server: the thing that lets an AI assistant ask your restaurant a question and get a real answer back. It is registered from the platform, so an agent finder, the index an assistant checks to see what is available, can pick it up.
Then a group where one person cannot eat gluten asks their assistant for somewhere that works — and gets your actual options instead of a review from two years ago. That question decides where four people eat, and it is asked before anyone opens a website.
- 1You describe what you sell
- 2It's published and registered
- 3A customer's assistant asks
- 4It answers in your words
- 5The details reach you
What AI can and cannot do here today
What AI can do today
- It can say which dishes suit which diets, from the allergen notes you have written.
- It can answer the questions that decide a booking — vegan options, kids menu, opening hours.
- It can point the guest to you to reserve while they are still choosing.
What it cannot do yet
- It cannot speak for the kitchen tonight. A substitution or a cross-contamination question goes to your team, not to a guess.
- It cannot read a menu you have not described. Change a dish and the notes need changing too — nothing is pulled from a live menu.
Everything on the left is what a text-based setup supports today. Quoting against live data, and taking the booking itself, need server types we hope to add later.
How the AI menu inquiry workflow works
- 1
Guest asks a question
A diner messages by WhatsApp, SMS, or chat about the menu, allergens, or hours.
- 2
AI answers from your menu
The agent works from the menu, allergen notes and opening hours you have described to find the right answer.
- 3
Accurate answer sent
The guest gets a clear, correct reply about options that suit them.
- 4
Handed over to book
The agent points the guest to you to reserve a table while they are still deciding.
Tools in this workflow
These are the tools the workflow touches. Map them in the Agentic Planner to describe what your agent can do.
WhatsApp Business
Send text, template, image, document, and audio messages to WhatsApp users via the Meta WhatsApp Business Cloud API.
Twilio
Send SMS and MMS messages, initiate and monitor phone calls, look up phone number metadata, and manage Twilio messaging resources from agent workflows.
Google Sheets
Integrate an AI agent with Google Sheets to read, write, and sync data — agents log results, look up reference tables, and update task status directly in your spreadsheets.
Industries this fits
Build this workflow
Open this exact workflow in the free Agentic Planner and adapt it to your business — no code required.
Related use cases
AI review response automation
Reply to every review quickly and on-brand — without it eating your day
Turn "book me a table" into a confirmed reservation
AI takes the request the moment a diner reaches out — no missed calls, no back-and-forth.
Turn catering inquiries into quoted bookings
AI qualifies every catering request and sends a tailored quote — without pulling you off the line.
Frequently asked questions
For anything it can't confirm safely, the agent hands off to your team rather than guessing.
Yes — the agent can answer wherever guests reach out, from WhatsApp to text.
Only from what you have written. Fugentic publishes your menu and allergen notes as an agent an assistant can ask, and registers it so an agent finder — the index an assistant checks to see what is available — can pick it up. The alternative is an assistant inferring your allergens from reviews, which is exactly what you do not want.
Make your services available to AI
Describe what you sell, what it costs and how you decide — then publish it as an agent an assistant can ask. No code.