For owners who have been pitched a chatbot

AI for Hospitality,Pointed at the P&L

Hospitality is one of the slowest industries to adopt AI — and it is not because owners are behind. It is because almost everything sold to them is a chatbot. A WhatsApp bot does not tell you which three dishes stopped being profitable when butter went up. This is what the other kind looks like, and how to adopt it without betting the venue on it.

14-day free trial
No credit card required
Setup in 15 mins

Why Hospitality Has Not Adopted AI Yet

Every owner has now been pitched "AI". Most of them tried something, found it did not change the month, and quietly stopped. That is a rational response to what is being sold — not resistance to technology.

  • What gets sold is a chatbot — a WhatsApp bot, a voice bot, a reservation replier. It answers guests. It never opens your books.
  • It is scored on the wrong thing: messages handled, not money kept. You cannot find either in your P&L at month end.
  • It arrives as one more subscription and one more login, disconnected from your POS, your stock and your roster — so it knows nothing about your venue.
  • It gives generic advice. "Optimise your menu" is not advice; it is a category. Which dish, at what price, from when — that is advice.
  • It asks for trust up front. Owners are asked to let software act on a business with 3% margins, on day one, with no track record.
  • Nobody explains what happened afterwards. No AI tool tells you whether its own suggestion actually worked.

What AI in Hospitality Should Actually Do

The useful version is unglamorous and it never talks to your guests. It reads what your venue already produces — orders, purchases, stock counts, rosters, bills — and does the work an analyst would do if you could afford one. It finds what changed, prices it, proposes the fix, and then checks whether the fix worked.

  • It watches the money, not the messages — cost, waste, margin, labour, quiet regulars, supplier prices.
  • Every finding carries a number, because "your food cost is high" is not actionable and "these three dishes now lose money, here is the new price" is.
  • It proposes a specific change you can preview before anything happens.
  • You decide how far it may go — advise you, ask you, or handle the small ones itself. The dial only moves when you move it.
  • It marks its own homework: weeks later it reports whether the change made the difference it predicted.
  • It runs on the venue you already have, because it lives in the same system as your menu, orders, stock and staff.

The Four Parts, and What Each One Is For

It finds what you cannot watch

Some things go wrong quietly, and they are the expensive ones. A set of checks runs across your own data on a schedule, looking for the specific ways hospitality leaks money.

  • Supplier prices creeping up without the menu following
  • What you bought versus what you sold — and the gap between
  • Discounts, comps and voids that stopped looking accidental
  • Regulars who quietly stopped coming
  • Findings arrive with the evidence attached, not as a score out of ten

It reasons about your venue specifically

A team of specialists — revenue, food cost, procurement, scheduling, guests and more — each looking at your numbers through one lens, and a briefing each morning that says what changed and what it is worth.

  • Twenty specialists, each with one job, not one assistant with a vague one
  • A morning briefing: what was handled, what needs you
  • Ask it anything about your venue in plain language and it queries your real data
  • It remembers your venue — your suppliers, your seasons, your decisions

You hold the dial

This is the part most AI skips, and it is the reason adoption stalls. Nothing acts on your business until you say it can, and permission is earned one step at a time rather than granted on day one.

  • Advise: it tells you, you do it
  • Approve: it prepares the change, you press the button
  • Autonomous: once it has a track record you trust, small changes go through on their own
  • Every action shows you a preview of exactly what will change, and is reversible
  • It can be moved back down the ladder at any time, by you or automatically

It tells you what to fix first

The consulting half. Answer questions about your venue — where you are, what you run, how you operate — and get an installable plan rather than a PDF: the standards, checklists and controls a good operator would put in, in the order they should go in.

  • Built around your country and venue type, not a generic template
  • Every recommendation can be installed into the system, not just read
  • Published playbooks written for real venues, not slideware
  • It tells you what is missing, and what it costs you to keep missing it

What You Are Actually Getting

20
Specialists
Revenue, food cost, procurement, scheduling, guests and more — each with one job
3
Levels of control
Advise, Approve, Autonomous — you set it, and it is earned, not assumed
Daily
Not quarterly
A briefing every morning, so problems are days old rather than months
$49
Per month
Locked for the first 100 venues. Less than one shift of the analyst it replaces

The Chatbot vs The System

What is usually sold as "hospitality AI"
  • Talks to guests; never opens your books
  • Measured in messages answered
  • A separate tool, blind to your stock, costs and roster
  • Generic advice that fits any venue, so it fits none
  • Wants full autonomy on day one, or offers none at all
  • Never reports whether its advice worked
  • One more subscription, one more login
FULLGUEST
What we built instead
  • Reads orders, purchases, stock, rosters and bills — the things that move margin
  • Measured in money found and money kept
  • Lives inside the system that already runs your venue
  • Names the dish, the price, the supplier, the shift — with the number attached
  • Permission is a dial you control, earned one rung at a time
  • Comes back weeks later and tells you if it actually worked
  • One system, one bill, one version of the truth

Questions Owners Ask Before They Trust Any of This

Why is hospitality so slow to adopt AI?+

Because the offer has been poor. Most hospitality AI is guest-facing chat — useful at the edges, invisible in the P&L. Owners run on thin margins and judge tools by whether the month changed. A tool that answers messages faster does not change the month, so it gets dropped. The blocker is relevance, not appetite.

Is this just a chatbot with a different name?+

No, and the simplest test is what it reads. A chatbot reads your messages. This reads your orders, purchases, stock counts, rosters and bills, and its output is a specific change to make with a number attached — reprice these three dishes, this supplier moved, Tuesdays are over-staffed by one.

Will it do things to my business without asking?+

Not unless you decide it can, and not on day one. It starts in advise-only mode. You can move it to asking permission per action, and later to handling small changes itself once it has built a record with you. Every action previews exactly what will change before it happens, and can be undone. You can move it back down at any time.

We are not a big group. Is this only for chains?+

The opposite. Groups can afford an analyst, a finance manager and a procurement lead. A single venue cannot, and that is exactly the gap this fills — the back-office roles you could never justify hiring.

Do I have to replace my POS to use it?+

It works best when it can see everything, because visibility is the whole point. That said, it plugs into common systems rather than demanding a rip-and-replace, and most venues start with one part — the menu, the stock, the money — and widen from there.

How do I know its advice is any good?+

Because it grades itself in public. When it proposes a change it also states what it expects to happen. Weeks later it reports what actually happened. Over a few months you are not trusting a claim, you are reading a track record.

What does it cost, and what is the risk if it does not work?+

The decision engine is $49 a month, locked for the first 100 venues, and there is a free digital menu you can start with at no cost and no card. The honest risk is not money, it is attention: any system like this is worthless if nobody reads what it says. Start with one thing you already suspect is leaking.

Where should a venue actually start with AI?+

Not with a chatbot. Start where you already suspect money is going and you cannot prove it — usually food cost or waste. Get that one thing measured properly for a month. A single real number changes how the whole team argues about decisions, and it is a far better first step than automating a conversation.

See it read a real venue

A live demo venue, with real findings on the board — the money it flagged, the changes it proposed, and whether they worked. Look at that before you believe any of the above.