Capability building
Pilot
This program grounds executive teams in generative AI for hotels, with hospitality-specific use cases, governance guidance, and facilitation-led working sessions.
Audience
Format
Duration
Category
§ 01Lab overview
This program grounds executive teams in generative AI for hotels.
Rather than treating AI as a trend presentation, the program combines guided facilitation, hospitality examples, and hands-on workflow mapping. Teams move from shared vocabulary into concrete pilot choices, governance patterns, and internal enablement actions. Six modules, run as a cohort or mixed and matched to a team's priorities.
Also available as a 2-day intensive for time-constrained cohorts.
§ 03Curriculum
6 modules · 24 hours · 6 half-day workshops
Six half-day, hands-on workshops that take hotel teams from AI fundamentals through prompting, research, data, content and prototyping. Every session is built on real hotel scenarios and produces something usable before the room empties.
Programme objectives
§ 04Client perspective
“The program gave our leadership team a practical language for AI adoption instead of abstract hype.”
“We left with an agenda for pilots, governance, and team enablement that felt realistic for our business.”
§ 05Ready to talk?
The Horwath team can adapt the format, the cohort mix, and the working agenda around the decision you actually need to make. Tell us what you’re working through and we’ll come back with a brief.
4 hours
Understand AI, set guardrails, enable adoption
Your team is already hearing about AI from every direction, but most of what they hear is hype. This workshop provides a grounded, hospitality-specific understanding of what AI can and can't do, where the real opportunities are, and how to adopt responsibly.
Learning objective
Develop shared understanding of AI capabilities in hospitality, governance frameworks, and adoption approaches.
Who it’s for
Leadership, department heads, all staff
You will learn to
What you take away
Large language models explained with hotel analogies, no maths required.
You leave with: A plain-language mental model of how AI works that you can explain to your own team.
Real vs overhyped AI applications with named hospitality case studies.
You leave with: A grounded read on which AI applications are delivering in hotels today — and which to ignore.
Classify your tasks (automate / augment / leave alone) and map AI opportunities across your property.
You leave with: An AI opportunity map of your own department, with every task classified automate / augment / leave alone.
Establish governance principles, risk guardrails, and a responsible adoption framework — plus where AI is heading.
You leave with: A governance principles checklist your property can adopt from day one.
4 hours
Communicate with AI effectively and consistently
The difference between useful AI output and wasted time comes down to how you communicate with it. This workshop teaches a structured prompting method that consistently produces high-quality results for real hotel tasks, from guest communications to operational analysis.
Learning objective
Construct structured prompts that consistently produce high-quality, usable outputs for daily hotel tasks.
Who it’s for
All staff
You will learn to
What you take away
Write a live prompt to a real guest complaint, then reveal the four-part operating brief and apply it to a task of your own.
You leave with: The four-part operating brief — Task · Context · Boundaries · Deliverable — applied to a task from your own week.
What to do when AI output misses the mark: the failure families, constraint design, and the draft–run–steer–bottle loop.
You leave with: The iteration discipline: name which of the four parts broke, and the move that fixes it.
The playbook system: bank the room's tasks, rank them, and bottle the first entry live.
You leave with: A started prompt playbook — Entry
4 hours
Smarter research, stronger institutional memory
Hotels depend on two forms of knowledge: reliable intelligence from outside the organisation and operational expertise held within it. AI can accelerate access to both — but only when teams know how to find the right evidence, verify what they receive, and ground AI in trusted hotel information. In this hands-on workshop, participants work in pairs to conduct AI-augmented research and build a working hotel knowledge assistant using accessible, off-the-shelf tools.
Learning objective
Use AI to conduct structured, evidence-backed research and create a document-grounded hotel knowledge assistant using accessible off-the-shelf tools.
Who it’s for
Management, strategy teams, operations leads, HR and training teams, and department heads responsible for research, organisational knowledge or operational decision-making.
You will learn to
What you take away
Scope → Gather → Curate → Synthesise → Verify → Deliver: a guided research assignment in pairs, turning a broad hotel question into a focused brief and a management-ready output — with every decision-critical claim verified against its evidence.
You leave with: A reusable research workflow, research-brief template, source-evaluation framework, claim-verification checklist and management-output structure.
Build and test a working knowledge assistant on a prepared hotel document store using no-code, off-the-shelf tools — then probe how it handles missing, conflicting and outdated information, and when a more advanced RAG solution is justified.
You leave with: First-hand experience building and testing a hotel knowledge assistant, plus a reusable document-store guide, assistant configuration template, testing checklist and RAG decision framework.
4 hours
Answers you can defend, and a report that rebuilds itself
Every hotel runs on spreadsheets: occupancy and RevPAR by month, channel performance, F&B covers, guest scores. Getting a decision out of one takes hours, and rebuilding the same report every month takes hours more. This workshop covers both jobs, individually, on a real hotel workbook or one from your own property: interrogating the numbers until you have an answer you can defend, then building the report that produces itself. Each half closes on what you can connect next, from the systems that feed the spreadsheet to the tools that do this work for you.
Learning objective
Use AI to interrogate a hotel spreadsheet and reach a defensible decision, automate the report you rebuild every month, and judge which data sources and tools are worth connecting next.
Who it’s for
Revenue managers, finance and operations leads, department heads, and anyone who builds or reads the same report every month.
You will learn to
What you take away
Watch a plain-language question about a hotel workbook produce a confident, well-formatted, wrong answer, then take it apart. Each stage is taught, demonstrated live on the same workbook, then run by each participant on a question of their own, through Frame, Prepare, Analyse, Check and Deliver. Closes on the tools that read a spreadsheet directly and the exports your PMS, RMS and POS will already give you.
You leave with: A verified one-page recommendation on a question from your own property, the five-stage workflow that produced it, and a shortlist of data worth pulling next.
Start from a finished, automated hotel report, then take it apart and build your own. Four stages, each taught and then run on a report you actually produce: Plan, Process, Present, and Schedule & Test, with the cleaning, joining and aggregating written as a plain-English recipe that AI runs every month. Closes on the ladder from an AI recipe to the same recipe in Excel or a database to a live connection, and on the tools that do this natively.
You leave with: An automated version of your most repetitive report, tested against a messy file, with handover notes so someone else can run it.
4 hours
Copy that sounds like you, and images you can publish
Hotels produce content constantly: offers, newsletters, menu descriptions, social posts, owner updates. AI drafts all of it in seconds, and a reader can spot the result at a glance, because it sounds like every other hotel using AI. This workshop fixes that from both ends: capturing your property's real voice from what you have already published, then briefing and steering images and video that look like your hotel rather than a stock library. Both halves end with something you can hand to the team, and a line about what AI must never produce in your name.
Learning objective
Capture your property's voice as a reusable anchor, produce real written and visual content with it, and set the boundaries on what AI may generate in your name.
Who it’s for
Anyone who writes something a guest or an owner will read: marketing and communications, sales, F&B and revenue writing offers, HR writing job ads, and whoever signs it off.
You will learn to
What you take away
Generate a piece of hotel copy with a plain prompt, then look at what the room recognises instantly: fluent, confident and anonymous. Each stage is taught, demonstrated live, then run by each pair on content they actually owe someone this week, through Capture, Anchor, Draft, Edit and Bank. Closes on where the anchor lives so the team uses it by default, and on the claims a human has to check before anything ships.
You leave with: A voice anchor drawn from your own published content, one finished piece written with it, and the edit checklist that got it there.
Start from a finished set of on-brand assets made live in the room, then take apart how each one was made. Five stages, each taught and then run on your own campaign: Brief, Generate, Steer, Check and Publish. Closes on where AI video and audio have actually got to, and on the line your property will not cross, from guest faces to rooms you do not have.
You leave with: A reusable visual brief, a set of assets for a real campaign, and a written do-not-generate line for your property.
4 hours
Build the tool, automate the job, no code
Every hotel has jobs nobody has time to fix: a form that is still paper, a report copied between three systems by hand, a request that sits in an inbox until somebody remembers it. Some of those need a small tool that does not exist yet. The rest need the systems you already pay for to talk to each other. This workshop does both, in pairs, on a problem from your own property, and neither half involves writing code.
Learning objective
Specify and build a working prototype for a real operational problem, automate a manual process across the tools you already use, and judge which of the two a problem actually needs.
Who it’s for
Operations managers, department heads, project leads, and the IT or systems owners who field these requests.
You will learn to
What you take away
Watch a vague request get built into the wrong thing, then trace the fault back to the specification rather than the build. Each stage is taught, demonstrated live, then run by each pair on a problem from their own property, through Frame, Spec, Build, Test and Decide. Test means handing your prototype to another pair, who use it wrong on purpose.
You leave with: A working prototype built from your own specification, the spec template that produced it, and a keep, bin or escalate decision.
Watch an automation fire live in the room, a review arriving, getting classified, drafted, routed and logged without anyone touching it, then take it apart. Five stages, each taught and then run on a process you currently do by hand: Map, Wire, Insert AI, Test and Hand Over. Closes on the ladder from Zapier and Make to a self-hosted n8n, and on what an automation must never do unattended.
You leave with: One real process automated end to end, the ladder for choosing a platform, and a written list of what stays under human control.
