A short course covering the vocabulary, risks and controls you need before your first pilot.
Key takeaways
Before evaluating any vendor, it helps to separate a tool that answers a booking or availability question from one that makes a judgment call, like waiving a fee or handling a serious complaint. Almost everything worth deploying at a restaurant or hotel falls into the first category.
This lesson gives operators a shared, plain-language vocabulary so vendor conversations don't get lost in marketing terms like 'AI concierge' or 'autonomous', which mean very different things depending on who's selling them.
The risk isn't that an AI assistant occasionally misunderstands a guest; every host stand does that too. It's the absence of a clear handoff to a manager when the conversation needs judgment, a refund, or a genuine apology.
This lesson walks through the common failure points, an unclear escalation path, a tone that doesn't match how your team actually talks to guests, so you know what to ask a vendor before signing anything.
Guest names, contact details and dietary or accessibility notes all count as personal information under Canadian privacy law. You need to know where that data goes once it enters a tool, how long it's kept, and whether your reservation or CRM data is used for anything beyond your own business.
This lesson gives a short checklist to bring into any vendor negotiation, so data handling is settled before a pilot starts, not renegotiated after a concern comes up.
The final lessons cover how to design a review step for anything guest-facing, who approves a response tone, how outbound messages get logged, and how to scope a first pilot narrow enough to judge fairly within a month or two.
By the end, you'll have a one-page framework you can apply to any AI pitch that lands on your desk, whether it's for reservations, reviews or the back office.
A 30-minute call is enough to tell you whether AI pays for itself here.
