AI training where every employee builds something they use the next day.
Most AI training is a demo. Someone shows the tools, everyone nods, and nothing changes. In my sessions each person brings one tedious task from their own week and leaves with a working AI system for it, built in the tool your company already pays for: ChatGPT, Copilot, Gemini or Claude.
Why most AI training does not stick
Companies buy seats, run a session, and six weeks later usage is back where it started. The cause tends to be one of the four below.
It teaches tools instead of tasks
A tour of features gives people nothing to do on Monday. Training sticks when each person works on a task they already do every week.
It happens once
One session raises awareness. Skills come from repeating the same workflow on real work until it is faster than the old way.
Nobody says what is safe to paste
Without a written rule, careful people avoid AI and careless people paste client data into free tools that may train on it.
Nobody measures it
If no one writes down how long a task took before and after, the time saved stays invisible and leadership stops paying for the program.
Every person, one real task, one working system
- 01
Score the week
Each person lists their recurring tasks and scores them on a 0 to 10 readiness scale: how much of the task AI can do well today, and how risky a mistake would be.
- 02
Pick one task
They choose the highest-scoring task that takes real time. Drafting, summarizing, first-pass analysis and reformatting usually win.
- 03
Build the system
A system is four parts: instructions, an example of good output, the input, and a check a person runs before anything leaves the building. Everyone builds theirs live, in the AI tool your company already pays for.
- 04
Run it on real work
They run it on this week's version of the task, compare it with the old way, and note the minutes. That number is the start of your ROI case.
- 05
Name the next three
Each person leaves with the next three tasks worth handing off, so the habit carries past the session.
Formats and fees
Up to 30 to 40 people building hands-on. Each session is tailored to your team's real work before the day.
90 minutes
$7,500
One build. Everyone scores their week, builds a system for one task and runs it before they leave.
Half-day
$15,000
Two builds, plus a first draft of your team's AI policy: what is safe to paste, which tools are approved, and who checks what.
Full day
$20,000
Everything in the half-day, with more time to build each person's systems and test them on live work before the day ends.
What your team leaves with
- A working AI system for one real task from each person's week
- Everyone's recurring tasks scored on the 0 to 10 readiness scale
- The four-part system recipe, reusable on the next task
- A one-page team AI policy starting point (half-day and full day)
- A before-and-after time count for the task each person built
Who it is for
Sales, marketing, operations, HR and professional-services teams, in any industry where the work runs on email, documents and meetings. Nobody needs a technical background.
I built the method running my own businesses on AI and teaching it to event teams. More than 2,000 people have used the custom GPTs I built, and the audience at Destination California rated my session 4.9 out of 5. For event teams specifically, see the event workshops.
Making it stick after the session
A session gets everyone started. Changing how a whole team works takes a few weeks of practice on real tasks, a written policy and a number to show leadership. If you run that yourself, these guides cover it: a four-week AI training plan, an AI policy template for small teams and how to measure AI ROI.
If you want it run for you, the 30-day AI pilot puts a small group on real work for a month and ends with a written plan for rolling it out to everyone.
Frequently asked questions
What makes AI training for employees work?
Build it around real tasks, not tools. Each person should leave having used AI on something they do every week, with a written rule for what is safe to paste and a way to measure the time saved. Spacing practice over a few weeks makes it stick; a single tool demo rarely changes how anyone works.
Which AI tool should we train employees on?
The one your company already pays for. I teach in ChatGPT, Microsoft Copilot, Google Gemini or Claude, because the method transfers between them. Use a paid business plan with training on your data turned off. Free consumer tiers are the wrong place for company work.
Do employees need technical skills?
No. If someone can write an email and edit a document, they can build the systems in this training. The sessions are designed for sales, marketing, operations, HR and professional-services teams, not engineers.
How long does corporate AI training take?
A 90-minute session gets everyone one working system. A half-day or full day adds a second build and a team AI policy. For a whole team to change how it works, plan on a few weeks of practice after the session; the four-week plan in my AI training guide lays it out.
How much does corporate AI training cost?
My sessions are $7,500 for 90 minutes, $15,000 for a half-day and $20,000 for a full day. Each one is tailored to your team's real work before the day.
How many people can attend?
Up to 30 to 40 people building hands-on. The room can scale to about 100 if some attendees watch rather than build.
Bring hands-on AI training to your team
Tell me who is in the room and what their weeks look like. I'll suggest the format and the tasks worth building first.