The AI grounding your team needs. Ending with something they built.
Everything the standard AI training covers, what these tools are, how to use them well, where they fail, and how to use them safely. Plus the part nobody else does: an hour where every person builds something real.
Most foundations courses teach the same five things. So do we.
We're not going to pretend otherwise. Any decent introduction to AI covers the vocabulary, the prompting, the use cases, the risks and the rules. Yours should too, that's what your team actually needs and it's what you'll be comparing us on.
The difference is the last ninety minutes. Most training days end with a slide deck and good intentions, and the industry's own honest estimate is that behaviour reverts within about a month. Knowledge decays. A working tool doesn't. So we end ours by having everyone build one, with their own hands.
Five modules, in order.
What these things actually are
How large language models work, in plain English, in about twenty minutes. Enough that your team stops treating it as magic and starts treating it as a tool with known behaviour. What it's good at, what it's structurally bad at, and why it makes things up.
Getting good answers
The practical craft. Context, specificity, giving it examples, telling it what "good" looks like, iterating instead of accepting the first response. Everyone leaves with a prompt library written for their own role, not a generic handout.
Finding the work worth automating
A structured session on your team's actual workflows. Where's the repetitive work, where's the judgement, and which of those should AI go anywhere near. Output is a prioritised list your team wrote themselves, which is the only kind anyone acts on.
Using it without putting the business at risk
What should never be pasted into a public AI tool. Where your data goes. Why the confident answer is sometimes the wrong one, and how to check. Client confidentiality, personal data, and the difference between a tool your organisation approved and one someone found on the internet.
Build something
Ninety minutes, hands on keyboards. Everyone makes one small working thing. Not a mock-up. A thing with a URL.
AI literacy you can evidence.
If your organisation deploys AI systems and operates in or serves the EU, Article 4 of the EU AI Act has required you to ensure a sufficient level of AI literacy among staff since February 2025. It applies to organisations using AI, not just those building it, which catches a great many businesses that assume it doesn't apply to them.
Notably, generic e-learning is unlikely to satisfy it: the wording requires training that accounts for people's actual roles, experience and context. A single module for everybody doesn't do that.
Every Foundations session includes a documented training record, who attended, what was covered, when, issued afterwards as a PDF for your files.
We're trainers, not lawyers, and we won't tell you this makes you compliant. What we'll tell you is that if someone asks what AI training you've provided, you'll have an answer with a date on it.
Half day or full day.
| Half day | Full day | |
|---|---|---|
| Length | ~4 hours | 09:00–17:00 |
| Modules | 01–02, 04, short build | All five, in full |
| Prompt library | Shared | Per role |
| Use case mapping | Group exercise | Per team, prioritised |
| Build time | 45 minutes | 90 minutes |
| Documented training record | ✓ | ✓ |
| From | £1,250 | £2,000 |
Up to 30 attendees. Plus VAT and travel. At your venue anywhere in the UK, or online.