# AI for hospitality — bookings, group enquiries, reviews, suppliers > What an agent takes on in a restaurant or a group: group and event enquiries, bookings, drafting replies to reviews, supplier orders. What it never validates: an allergen, a public reply. **Version HTML :** https://agence-intelligence-artificielle.eu/en/ai-hospitality **Langue :** en-GB --- AI by industry # AI for hospitality An enquiry for thirty people arrives at 8.15 pm, in the middle of service. It is read at midnight, and the proposal goes out two days later — when the customer has already booked elsewhere. The task that costs in hospitality is not in the kitchen or on the floor. It is everything that arrives in writing while nobody has a free hand. [Request the survey](https://agence-intelligence-artificielle.eu/en/contact) - **5 tasks** documented below - **1 limit** no allergen validated, no review reply published without review - **No fake reviews** never, neither written nor misleadingly solicited - **2 to 6 weeks** before the first system is open The cost ## What this industry pays when nothing moves. Ranges, not promises. They come from tasks we meet in this industry and orders of magnitude we measure in the survey — yours may fall outside, in either direction. **Changing nothing** - Time to answer a group enquiry : **1 to 3 days** - Group enquiries unanswered for lack of time : **1 in 4 to 1 in 6** - Online reviews with a reply : **1 in 3, when someone has time** - Time per group proposal : **30 to 60 minutes** - Supplier orders placed from memory : **often, with omissions** - Groups lost to slowness, per month : **2 to 6** **The enquiries system, opened** - Time to answer a group enquiry : **acknowledged within a minute, proposal within hours** - Group enquiries unanswered for lack of time : **fewer than 1 in 20** - Online reviews with a reply : **all, drafted the same day, published after review** - Time per group proposal : **5 to 10 minutes of review** - Supplier orders placed from memory : **prepared from stock levels, validated** - Groups lost to slowness, per month : **0 to 1** Five tasks ## What we remove, and what we leave. Each task is described with what the agent takes on and what stays with a person. If the second column is empty, we have not understood the industry. 01 ### Answer group and event enquiries **What the agent does** It acknowledges within a minute, extracts date, numbers, package, constraints, checks the schedule, and pre-fills the proposal from your packages. **What stays with a person** The final availability, the price, the exceptions. A room promised twice costs more than three hours of delay. 02 ### Handle bookings and reminders **What the agent does** It confirms, sends a reminder the day before, offers another slot when full, and keeps the waiting list. **What stays with a person** Seating, whether to give the table or not, the exception for a regular. The floor decides. 03 ### Draft replies to reviews **What the agent does** It drafts a reply to every review, in the house tone, taking into account what happened that evening if you noted it. **What stays with a person** Publication. Nothing goes out without review, and no review is ever written or solicited in a misleading way. 04 ### Prepare supplier orders **What the agent does** From recorded stock levels, the week’s menu and history, it prepares the order per supplier, with deviations from the usual. **What stays with a person** The validation. An order placed without the chef is a cold room full of what you did not plan. 05 ### Keep allergen sheets and menus up to date **What the agent does** When a recipe changes, it prepares the update of allergen sheets and menus, and flags what is not filled in. **What stays with a person** The validation of every allergen by the chef. A forgotten allergen is not a data entry error. What we do not do ## No allergen is validated, no public reply is published by an agent. The two places where a mistake cannot be undone. The agent prepares sheets and replies ; the chef validates allergens, a person reviews what becomes public. And never, in any case, is a review written or solicited in a misleading way. The diagram below shows the circuit of a group enquiry. **The method, in one sentence** Ten systems, one survey, one system opened at a time — the one with the shortest payback. [Read the full method](https://agence-intelligence-artificielle.eu/en/#methode), or go back to the [AI agency](https://agence-intelligence-artificielle.eu/en/) page. The tools ## What we open it with, and why here. The choice of tool comes last, and it is reversible. But in this industry, three constraints keep coming back and steer the choice before we even start. **Your booking tool, as it is** OpenTable, ResDiary, SevenRooms or a shared diary : the agent reads what it exposes and writes into what it accepts. It replaces nothing. **A circuit visible in n8n or Make** Packages change with the season. A rule is corrected in minutes by you, without calling us. **A model that keeps your tone** It learns from your past replies and reproduces it. You review — and correct, often, in the first weeks. Then hardly at all. The objections ## What we hear in this industry, before we start. ### We reply to reviews ourselves, it is our voice. That is why it drafts and does not publish. It removes the blank page and the delay ; the voice stays yours, because you review and you change it. A review waiting three weeks for a reply — that is not your voice. ### Allergens are too serious for a machine. Yes, and it is written into the quote : it prepares the sheet, the chef validates every line. What it removes is the forgotten update when a recipe changes — the place where the error is born. ### How long before it is useful ? The survey of the ten systems is delivered within 72 hours. Opening the first system takes two to six weeks. Ideally off-season, so it runs when you no longer have time to think about it. ### A single restaurant — is this for us ? A single task — group enquiries — is enough to change the week of a restaurant that does functions. For a group of restaurants, we open the same system across every site at once. Where this industry concentrates ## The areas where we are called most for this. [**London**AI agency in London](https://agence-intelligence-artificielle.eu/en/ai-agency-london)[**Manchester**AI agency in Manchester](https://agence-intelligence-artificielle.eu/en/ai-agency-manchester)[**Edinburgh**AI agency in Edinburgh](https://agence-intelligence-artificielle.eu/en/ai-agency-edinburgh) [**All locations**see the full map](https://agence-intelligence-artificielle.eu/en/locations) Related industries The same tasks, in neighbouring industries : [AI for e-commerce](https://agence-intelligence-artificielle.eu/en/ai-ecommerce), [AI for sales teams](https://agence-intelligence-artificielle.eu/en/ai-sales), [AI for real estate](https://agence-intelligence-artificielle.eu/en/ai-real-estate). First call — fifteen minutes ## We will tell you which system to open first. Describe the task that costs you the most. We will tell you what is feasible, how long it takes and what it costs. If the answer is no, you will leave with the two reasons why. [Request the survey](https://agence-intelligence-artificielle.eu/en/contact) [Write to the agency](https://agence-intelligence-artificielle.eu/en/contact) --- ## Autres langues - fr-FR : https://agence-intelligence-artificielle.eu/ia-restaurant.md - en-GB : https://agence-intelligence-artificielle.eu/en/ai-hospitality.md - de-DE : https://agence-intelligence-artificielle.eu/de/ki-gastronomie.md - es-ES : https://agence-intelligence-artificielle.eu/es/ia-restauracion.md - it-IT : https://agence-intelligence-artificielle.eu/it/ai-ristorazione.md - pt-PT : https://agence-intelligence-artificielle.eu/pt/ia-restauracao.md