Training a hospitality group's leaders and AI champions to use Claude and Gemini

Tom&Co AI recently ran AI training for a hospitality group with over 75 sites. Twenty-six people took part across two groups: the senior leadership team, and a group of AI champions from across the business, including Finance, Marketing, Operations and People. Each group had two sessions. The first was to understand AI, and the second was to build with it.

We started by asking

Before the first session, everyone filled in a short survey. Among the leaders, average confidence in using AI for their own work was 2.8 out of 5. Seven of the eleven had had no training. Nearly all had tried it for drafting or research, and their biggest worry was answers that sound right but are wrong. So we built the sessions around that worry.

Session one: what it is, what it's good at, and the rules

We covered how these tools work in plain English, where they're strong and where they let you down, and how to check an answer before you rely on it. We went through the company's AI policy before any demos. People get asked "am I allowed to put this in?" from day one, so they need the answer first.

Everyone then set up on their own accounts and watched live demos using the kind of work they actually do. The training covered Claude and Google Gemini, the AI built into the Google tools many of them already use.

For leaders: how to roll it out

Leaders don't just need to use AI themselves. They decide how it spreads. So the leadership sessions also coached them on rolling AI out across the company, in the same way you'd introduce any other new system: why you're doing it, who owns it, what the rules are, and how you'll know it's working. We also covered what only leaders can do, such as setting expectations and making it clear that AI supports people's work, not replaces it.

For champions: being the person colleagues ask

The champions' sessions covered the role itself. Colleagues will bring them questions and answers that look right but aren't, so we focused on checking, and on knowing where the policy draws the line.

Session two: everyone builds something

This was hands-on, using their own information in Claude. Each person worked through four exercises:

1. A writing assistant that sounds like them.

2. A weekly briefing that pulls from their email, calendar and files.

3. A presentation made from that briefing.

4. A 30-minute challenge to automate something from their own job.

People built things they could use the following week. One person made a Skill, a reusable set of instructions for Claude, that automates expense submissions. Another built a tool that checks how shifts have been allocated on the rota.

What worked

• Ask people what they know and worry about before you plan anything.

• Cover the rules before the demos.

• Use their work, not examples. People remember what they built.

• Teach a simple habit for every task: say what you want, give the background, check the result.

AI training and enablement for your team

Tom&Co AI trains and supports teams to use AI safely on their own work. That covers leadership sessions, champion programmes and hands-on workshops, built around your policy, your tools and your people, with coaching on how to roll it out across the business.

If you'd like to talk about training for your team, get in touch.