Insights
Ten questions to ask any AI trainer before you hire them
Ten questions to ask any AI trainer before you hire them, so you can tell who has actually run the workflows they teach and who is reading from slides.
Short answer: ask what happens minute by minute in the room, what your team walks out holding, whether they build your workflows or teach you to build your own, what happens to your data, and what support exists after the day. The ten questions below separate trainers who run these workflows daily from trainers reading slides.
AI training is a young market and the standard is uneven. Some trainers have run the workflow they are teaching a hundred times. Some have never opened the tool they are demoing outside of the slide they built about it. The difference is not obvious in a sales call, and it is very obvious three weeks after the workshop when nothing has changed.
We wrote this because clients keep asking us how to compare us with the other quotes on their desk. These are the ten questions we would ask if we were on the buying side. Ask them of us too.
1. What actually happens in the workshop, minute by minute
Listen for a rundown that includes the tool open on a real machine, not a deck of screenshots. A good answer names the tool, names the exercise, and names the deliverable the room walks out with. A weak answer talks about frameworks and journeys.
2. What does my team walk out with in their hands
Listen for artefacts. A prompt library saved into their account. A working workflow they built during the day. A written procedure they will use next Monday. If the answer is a certificate and a PDF of slides, keep looking.
4. What happens to my data during and after the session
Listen for a specific answer about which tool, on which plan, with which data handling terms. A good trainer knows whether your prompts are used to train the model on the tier they are demoing, and can say so without checking notes. A vague answer here is the biggest red flag on this list.
5. What happens after the workshop
Listen for a plan, not a promise. A follow-up call in two weeks. A Slack or email channel for questions. A 30, 60, 90 day review. AI training does not stick from a single day in a room, and any trainer who suggests it does is either new or is quietly hoping you do not ask.
7. How do you handle a room with mixed skill levels
Listen for a real method. Paired exercises so a confident person sits next to a cautious one. A parking lot for advanced questions. Optional stretch tasks for the fast finishers. A trainer who says everyone gets the same content is either underprepared or has not run enough rooms to know that never works.
8. What is your plan for the person in the room who does not want to be there
Every team has one. A good trainer expects them, welcomes them, and has a way to bring them along without a fight. Listen for empathy and specifics. A weak answer treats resistance as a problem to be steamrolled. A strong answer treats it as a signal about what the tool needs to do for that person specifically.
10. What is included in the quoted cost, and what is not
Listen for a specific list. Number of participants. Session length. Pre-work. Follow-up. Materials. Travel. GST. Whether recordings are included and whether they are yours to keep. Trainers who leave this deliberately vague are hoping you do not ask before signing.
How to use this list
Do not ambush a trainer with all ten in a single meeting. Pick three or four that matter most to you and ask them plainly. The point is not to catch anyone out. It is to make sure the person you hire is the person who will actually show up and do the work.
If you want the list as a printable PDF to take into the next conversation, email toby@ai-guy.co and we will send it across.
Related reading
For an idea of the workflows a good trainer should be able to demonstrate live, see how to run your inbox with Claude and how to automate repeat work with Claude.
If you want to put these questions to me, have a look at how I work.
