AI and Copilot Training

First, the distinction that saves you a wasted day

“AI training” covers two completely different subjects that share a name.

One is using AI tools: Copilot in Word and Excel, writing a prompt that gets a useful answer, knowing what the tool is likely to get wrong and how you would notice. No technical background needed, and this is what most organizations actually want when they ask.

The other is building AI systems: services, models, APIs, the architecture around them. Real prerequisites, and a developer or data audience.

People routinely book the second when they wanted the first, because the titles do not make the difference obvious. So work out which one you are before you look at any course.

If you want to use AI in your working day

You are the largest group and the most poorly served, because most of what is published is either marketing or aimed at developers. What you want is a practical day: how to ask for something and get it, where the tools are reliable, where they quietly are not, and what your organization’s rules mean in practice.

Take this before any certification course. Understanding what these tools do is what makes the rest make sense, and for a lot of people it is the only class they ever need.

If you need to understand AI without building it

Managers, analysts, project leads and anybody who has to make a decision about AI rather than write code. You need the concepts and the vocabulary, and you need to know what these systems can and cannot do well enough to spot a claim that is not credible.

It is a Microsoft certification and it is Azure flavored, but the ideas are general and the certification is recognized. If you want the understanding without the exam, tell us and we will say whether it is still the right class.

If you are building it

Different subject entirely, and real prerequisites: you should be a working developer, comfortable with an API and with at least one programming language, before you sit down.

The second one is there for a reason. A large share of stalled AI projects are stalled on data rather than on models, and the team usually knows it before anybody says so out loud.

If you are rolling it out for an organization

Then your problem is not which class to take. It is that Copilot licenses have been bought, a handful of people are using them well, most are not using them at all, and nobody can say what the difference is.

That is a training and governance question rather than a product one: what people are allowed to put into these tools, what has to be checked before it leaves the building, and who to ask. It usually wants a private session built around your own policies rather than a public class built around a sample tenant.

Tell us what you are rolling out →

The short version

• Have Copilot, do not get itthe Copilot class
• Deciding about AI, not building itAI Fundamentals
• Developer, building on Azurethe AI Engineer course
• Stuck on the datathe data class, not an AI one
• Responsible for the rollouttalk to us about private training

Training a team

This is the subject where private training makes the most difference, because so much of the value is specific to you. What tools you have licensed, what your policy actually says, what your people are already doing with these tools whether or not anybody approved it. A public class cannot address any of that.

We deliver online or at your offices, six people for on site and four for virtual.

How our classes are delivered covers what a seat includes and how organizations buy them.

Frequently asked questions