Google Gemini AI in Your EPOS: What It Actually Does Day to Day

A bit of context first

HDS Cloud Apps is the only hospitality EPOS platform with Google Gemini AI embedded directly at its data core. That is not a bolt-on integration or a third-party chatbot dropped in front of a dashboard. Gemini has access to your live transaction data, your menu, your stock levels, your staffing patterns and your historical trading — and it uses all of that to answer questions and surface insights without you needing to build a query, open a spreadsheet or wait for a monthly report.

What follows is a walkthrough of how that plays out in a real operation, from the moment service starts to the point the owner is reviewing the week's numbers on a Sunday morning.

What the chef sees at the start of service

Before a restaurant opens, the kitchen team needs to know what is likely to sell, what they need to prep, and whether anything on the menu is running low. Historically that information came from experience and gut feel. With Gemini built into the HDS platform, it comes from data.

The AI surfaces a pre-service digest: which dishes sold most in the same slot last week, what the weather is doing (a genuine factor in cover numbers and dish preference), whether a local event or school holiday is likely to affect footfall, and whether any ingredients are approaching a reorder point based on stock control records. A head chef can glance at this on a tablet and make prep decisions with confidence rather than instinct alone.

For restaurant operators, this kind of intelligence was previously available only to large groups with dedicated data teams. On HDS Cloud Apps it is part of the standard platform.

What the server sees mid-service

This is where Gemini's upselling capability earns its place on the floor. When a server is taking an order, the system analyses what is already on the table and suggests additions that are genuinely relevant — not a generic "would you like a dessert?" prompt, but a specific recommendation based on what the current table has ordered, what is proving popular with similar orders today, and what has the best margin.

A table that has ordered a sharing starter and two main courses might be prompted to consider a particular side that goes well with one of the mains. A table that ordered a bottle of wine early might be prompted at a logical moment. The suggestions appear on the Gemini 1 EPOS terminal as a brief prompt the server can act on or ignore — it is a nudge, not a script.

The same logic applies in bars and pubs. A customer ordering a craft beer round at the bar might trigger a suggestion for a bar snack that sells well alongside it. The system learns from what actually converts, so the suggestions improve over time as your trading data accumulates.

What the owner asks on a Monday morning

Conversational reporting is one of the most practically useful things Gemini brings to an EPOS platform. Instead of navigating a reporting interface and building a query, an owner can type a plain-English question and get an answer drawn from live data.

Common questions from operators include:

• "What was our busiest hour last Friday and how did it compare to the week before?"
• "Which tables turned the most covers last week?"
• "What is our gross profit on the new burger since we changed the supplier?"
• "How many covers did we do between 7pm and 9pm across the last four Saturdays?"
• "Which staff member had the highest average transaction value this month?"

Each of those questions can be typed into the HDS back office and answered in seconds. For operators running cafes, hotels or takeaways — where the owner is often also the manager and the accountant — this removes a significant administrative overhead from the week.

Scheduled reports that arrive without being requested

Not every insight needs to be requested. The HDS Gemini integration can be configured to send automated reports on a schedule — a daily trading summary by email, a weekly GP analysis, a Monday morning comparison of the previous seven days against the same period last year. These arrive without the operator having to remember to pull them.

For multi-site operators, this is particularly valuable. A group running sites in Manchester, London and Edinburgh can receive a consolidated morning summary across all venues, with variances flagged automatically. No logging into three separate back-office systems. No compiling figures in a spreadsheet. The AI does the aggregation and surfaces what matters.

The same capability is available to operators in Dublin, Glasgow, Birmingham and every other city on the HDS coverage map.

What it does for stock and margin

Stock management is one of the highest-value applications of AI in hospitality because the losses are so predictable. Wastage, over-ordering, portion drift and supplier price changes all erode margin silently — and they are very hard to catch without someone watching the numbers closely every day.

Gemini monitors your procurement and stock control data continuously. It flags when actual stock usage deviates significantly from what the sales data would predict — a signal that something is being wasted, pilfered or portioned inconsistently. It alerts you when a supplier's unit price has changed relative to your historical spend. It identifies which dishes are carrying a GP percentage below your target and suggests either a price adjustment or a recipe review.

For golf clubs and other operations with complex retail and catering mixes, this kind of margin intelligence across different product categories is genuinely difficult to maintain manually. The AI makes it continuous rather than periodic.

What it does not do

It is worth being clear about what the AI is not. It does not make decisions — it informs them. It does not replace experienced floor staff or a skilled head chef. It does not generate insight from data it does not have, which means the quality of what it surfaces improves as your trading history on the platform grows.

The AI is also not a separate product you need to manage. It is embedded in the same platform that runs your EPOS, payments, loyalty, kitchen display screens and back office. There is no separate login and no integration to maintain. You will need a Google Workspace account to access the Gemini features, but beyond that it works from the same data your team uses every day.

How to see it working in your business

The most effective way to understand what the Gemini integration will do for your specific operation is to see it against your own data. Our team runs demos that use anonymised real trading patterns from businesses similar to yours — restaurants, bars, cafes, hotels and takeaways — so you can see the kind of questions it answers and the kind of suggestions it makes in a context that makes sense for your team.

You can also explore the full list of HDS Cloud Apps modules to understand how the AI sits alongside the rest of the platform — from Musqet integrated payments to the Rush Hour demand forecasting tool — and how they work together as a single connected system rather than a collection of separate tools.