We ran an all-employee AI Month and now everyone in our company is ready to build

Like many tech companies, we had plenty of enthusiastic AI users in our business, and we'd hedged our AI bets by investing in Microsoft 365 Copilot and Claude Pro licences for the whole company. We knew there were varying levels of capabilities and confidence throughout the business. That's why it was important for us to make sure people were fully using the tools, as well as ensuring dollars going in meant productivity going up. So this August we ran "AoFrio AI Month", four weeks of company-wide sessions and workshops on using AI properly, run almost entirely by our own people. It was a massive undertaking, but we reckon the payoff for our business was worth the blood, sweat and tears. Here's how we did it, and what we learnt. 

1) One demo won't move anyone. A month might.

We knew that handing someone a powerful AI tool with no runway to learn it would set us up to fail. The last thing we wanted to do was put on a single all-hands training session and call it done. 

Over four weeks, we slowly immersed the team, starting with interactive and thorough walkthroughs with both Claude and CoPilot to build a baseline understanding of what each tool could do. The next two weeks were all about building. We held several hands-on sessions, tailored for different functions and needs, where people learned to solve real problems with their peers. The trainers were, for the most part, our own people. Many of our AI Advocates (a small representative group of employees helping to promote AI internally) stepped in as trainers and presenters, an excellent way to build advocacy within teams and create training examples matched to the real needs of our organisation.

This approach saw even the sceptics engage. Our biggest single session pulled in over 80% of our local employees in person, with more online. Even colleagues who'd told us earlier in the year they'd probably skip this type of event because they couldn't see how AI applied to their job showed up. By week four, several of them had built their own tools and agents and folded them straight into daily work. 

2) Everyone got their hands dirty, from total beginners to the technical crowd

The real shift we hoped AI Month would make was moving people past using AI to draft emails or summarise meeting minutes and start building things that could truly take work off their plate. 

And because nobody was told which tool to use, people worked out for themselves whether Copilot or Claude fit their needs better, rather than waiting for a recommendation from IT.

Colleagues who'd never built in AI before were creating agents and skills for time consuming admin tasks, like building on-brand presentations, writing job descriptions and processing expense claims. But even our more AI savvy colleagues got something out of the sessions. One of our most technically minded engineers walked away with new perspectives and a new working solution, built off ideas and tools he hadn't tried before. Beginner or expert, everyone ended up building, not just taking notes.

3) We closed the month with a showcase, not just a wrap-up email

We didn't let AI Month fizzle out on a Friday afternoon. We closed it with a company-wide showcase, with people demoing what they'd built, followed by a prize giving to call out the standouts. 

The prizes were a great incentive, but the real win was the pipeline of brilliant ideas we surfaced up. We now had over 50 standout use cases the business can go and build out properly. A few of the best concepts:

• A Supply Chain colleague who'd never touched an AI tool before designed an automated workflow that retrieves supplier documents, extracts relevant information, and matches it against purchase orders. This dramatically slashes manual effort and improved accuracy for our Supply Chain team.
• A cooler data interpretation tool, that could investigate connectivity problems on behalf of any employee within our internal tools. This was a bottleneck task that previously took engineers up to 3 hours to resolve.
• An agent that could pull CRM, support and product data into one comprehensive customer health view, something that used to require a lot of moving between multiple systems and lengthy reporting time to produce.

The bottom line

We set out to close the gap between occasional, uncertain use and people living and building in these tools every day, and by the end of four weeks that's exactly what happened. Confidence didn't just lift in the technical corners of the business. People in Supply Chain, People & Culture and Finance were building AI agents for themselves that a month earlier they'd never have thought possible. Perhaps most significant was the mindset shift. People changed from using AI just to assist with their own work to solving macro problems that impacted all parts of the business.

We crowdsourced over 50 agent and automation ideas from staff over the course of the month, with many of those built and being used by the time we closed the programme out. The bulk of these ideas could be mapped back to strategic organisation initiatives as well, an important outcome that really helped our leadership team see the value.

A month-long programme of this kind is something we'd recommend to any business that wants to rapidly ramp up real AI adoption across the organisation. 

We're already planning round two for next year to keep the momentum going. If you're a similar business weighing up whether a dedicated "AI Month" is worth the effort for your organisation, we're happy to swap notes on what we'll be doing differently and share our programme.