← All Sectors/Hotels & Restaurants
CASE STUDY · 29-ROOM RESORT

The Whole Resort in One System: Front Desk, Restaurant, Spa, Shifts and Stock

A resort with 29 rooms and 72 beds, a restaurant, spa and halls for events of up to 150 people. One system on top of the PMS and POS: guest replies, event proposals, restaurant and spa upsell, shift planning, stock with run-out prediction, and an overview of table service. Staff always send and approve.

3 SECONDS
Target: guest reply drafted in their language in 3 seconds, front desk approves in one click
Payback calculated from your own data in the pilot
Payback
Live at a 29-room resort
Status
Hotels & Restaurants
Target Industry
Resort Operations Suite · AerisQ
Open full app (8 modules) ↗
Weddings & Events
1 proposalright after the inquiry
Resort Size
29 rooms72 beds
Response Time (PMS)
3 sekundy5 languages (EN, CS, DE, PL, IT)
Event Halls
up to 150guests per event
Guest message · DE

Guten Tag, können wir am Freitag schon gegen 11 Uhr einchecken? Wir reisen mit unserem Hund.

Translation: Inquiry received at 10:47 pm after the front desk left. AI verified live PMS availability in 3 seconds; front desk sends with one click.

Drafting a reply
Guest
Familie BeckerBooking.com · DE
Received
10:47 pm · After hours
Booking
BK-48211 · room 214
PMS Check (3s)
Room free from 10 am (OK)
PMS Price list
Dog 300 CZK per night
The Problem: A Small Front Desk and Many Different Requests

In a smaller resort everything rests on a few people. The front desk answers guests and builds event proposals, the kitchen orders by feel, shifts are planned in a spreadsheet, and nobody knows exactly which restaurant tables earn the most or where service falls behind. The data sits in the PMS and POS, nobody just puts it together.

×The front desk answers guests and builds wedding and event proposals by hand
×Restaurant and spa are offered only on arrival, if at all
×Shifts are planned by guesswork, not by occupancy and events
×Ingredients run out unexpectedly or get ordered twice
×No overview of the busiest tables or how fast guests are served
The Solution with AerisQ: Assistive AI Connected to the PMS and POS

Assistive AI connected to the PMS and POS joins the data the resort already has. It drafts replies and proposals, offers the restaurant and spa at the right moment, suggests shifts from occupancy, watches stock and predicts what runs out when, and shows how the restaurant is served. A person always decides and sends.

A guest reply draft within 3 s and a first event proposal
Automatic restaurant and spa upsell before arrival and during the stay
Shift suggestions from room, restaurant and event occupancy
Stock: what was bought, what is missing and when it runs out
Table service analytics: occupancy, service speed and turnover
02 · TECHNICAL ARCHITECTURE

Step-by-Step System Flow

MODULE 01FRONT DESK & UPSELL

Front Desk, Guest Replies & Upsell

Messages from Booking, email and WhatsApp land in one place. Assistive AI checks PMS dates, drafts a reply in the guest's language, and builds event proposals. It also suggests free dining and spa slots before arrival and during the stay.

  • Reply in 5 languages within 3 s
  • Event, dining & spa proposals
  • Written back to PMS
MODULE 02SHIFTS

Shift Planning

From room occupancy, table bookings and planned events the system suggests how many people are needed at the front desk, in the kitchen, on the floor and in housekeeping. The manager only adjusts and confirms.

  • Based on occupancy and events
  • Front desk, kitchen, floor and housekeeping
  • The manager confirms the roster
MODULE 03STOCK & PREDICTION

Stock: What Was Bought and What Runs Out When

The chef photographs a delivery note and the stock receipt is waiting. The system compares purchases with POS sales, shows what is missing and predicts from occupancy and planned events when each ingredient runs out.

  • Stock receipt from a delivery-note photo
  • What was bought and what is missing
  • Prediction of what runs out when
MODULE 04RESTAURANT

Table Service Analytics

From the POS the system reads which tables are busiest, how long guests wait for ordering and for the bill, and how quickly a table is seated again. Management sees where service falls behind and where seats can be added.

  • Occupancy of each table
  • Service speed and waiting for the bill
  • Table turnover by time of day

We integrate directly with your existing PMS (HotelTime, Previo, Mews) and POS. We assess which module to deploy first during the 20-minute consultation and verify it in a pilot on your live operations.

03 · BEFORE AND NOW

What Changes for the Resort

How much time and money it saves you we measure in the pilot on your operation.

Front desk
BeforeReplies and event proposals written by hand, often the next day
With AerisQReply draft in 3 s and a first event proposal right after the inquiry
Restaurant and spa
BeforeGuests learn about free slots only on arrival
With AerisQAutomatic offers before arrival and during the stay
Shifts
BeforeRoster in a spreadsheet by guesswork
With AerisQShift suggestions from occupancy and events, the manager confirms
Stock
BeforeIngredients run out unexpectedly, notes retyped after closing
With AerisQReceipt from a photo and a prediction of what runs out when
Table service
BeforeNobody knows which tables earn the most or where guests wait
With AerisQOverview of table occupancy, service speed and turnover
04 · TARGET PARAMETERS

Modelled values; actual impact is measured on your data in the pilot

3 s
Guest reply draft
Target: in 5 languages (EN, CS, DE, PL, IT), checked in the PMS, approved in one click
4 modules
One system for the whole resort
Front desk & upsell, shifts, stock and restaurant on top of the PMS and POS
29 rooms
Resort size
72 beds, restaurant, spa and halls for events of up to 150 people
Prediction
What runs out in stock and when
Target: from occupancy, events and POS sales
TECH STACK:
Hotel PMS API (HotelTime, Previo, Mews)Dining & Retail POSMultilingual LLM (5 languages: EN, CS, DE, PL, IT)Mobile Vision OCR (delivery notes)Roster & Capacity Engine
05 · FREQUENTLY ASKED QUESTIONS

Everything to know before the pilot

Why not simply deploy an off-the-shelf chatbot from a no-code platform?+

An off-the-shelf chatbot can easily promise guests discounts or confirm rooms that are not available. Our assistant works from your live PMS data and hotel rules, presenting reply drafts for one-click staff approval. Nothing reaches a guest without approval, keeping communication 100% under your control.

Can the system handle wedding and corporate event inquiries?+

Yes. It reads headcount, date and wishes from the inquiry, checks a free hall and rooms and drafts a first proposal with menu and price. A person always confirms the final price and sends it.

Do we need to replace our current PMS or POS?+

Not at all. We build AI agents as a bridge directly on top of your existing hotel PMS and POS systems via standard APIs.

How does delivery note OCR work in practice?+

Staff simply snaps a photo of the paper delivery note on a phone. The OCR model extracts all line items and updates inventory within 5 seconds.

Is guest data compliant with GDPR?+

We process guest data in line with GDPR and it is never used to train public models. Zero data retention for cloud models is set up as agreed, or the system can run on-premise.

VERIFY FEASIBILITY IN YOUR OPERATIONS

Want to deploy a similar system in your company?

In a 20-minute consultation, we review 1 of your specific processes and tell you upfront whether and for how much it makes sense to automate.

06 · OTHER SHOWCASES

What we can build in other sectors

All Projects →