How do you choose AI training that your team can use at work?

Define who the training is for and how it should help

"We want to learn about AI" is a good conversation starter, but too broad a goal for evaluating a programme. Ask participants what tasks take up their time. Preparing a proposal, organising notes, and reviewing a summary require different exercises and evaluation criteria.

Gather information about experience. Someone who has never worked with such a tool will need time to learn the basics. Someone using it daily may expect more challenging examples. If levels differ significantly, discuss separate groups or a joint introduction with different exercises.

For example: “The participant will draft a response using our service information and check that it contains no unapproved commitments.” That outcome can be assessed directly. “Learn about the latest trends” is harder to apply to everyday work.

Assess the programme through the tasks participants will do

A list of tools can look impressive, yet convey little value in the session. Try adding a verb to each item: the participant will prepare, compare, improve, or verify. If it's all about watching a presentation, ask about the practical part.

Look for exercises in writing prompts, selecting materials and checking responses. Fluent wording does not mean the task is complete. Participants should recognise claims unsupported by the source and improve the response without losing important requirements.

Ask for an example that's similar to the team's work. It doesn't have to use real client data. A fictional, well-prepared case allows you to practise the method without accidentally revealing information.

Sample exercise: responding to an enquiry

Prepare a fictional service description with its scope, terms and details that have not yet been agreed. Add a customer enquiry that asks about an unconfirmed deadline. The participant should produce a useful draft response without inventing that deadline.

After the first attempt, the participant compares the response with the source material. They identify confirmed facts, unsupported promises and questions that need clarification. They then revise the prompt and check the next response.

Finally, introduce another enquiry to see whether the participant can apply the method independently, rather than just repeat the demonstration. This is a suggested exercise, not a report of a completed training session.

• Context: the recipient and purpose of the message.
• Materials: service information, terms and clearly identified gaps.
• Prompt: task, format and rules for using sources.
• Assessment: accuracy against the source, completeness and tone.

What can you practise in a 4-hour AI workshop?

Four hours provides time to focus on selected tasks and move from an initial prompt to a checked result. The scope must suit the group’s experience. Beginners need time to practise independently, make corrections and ask questions. A long list of tools should not crowd out those activities.

A sample set of tasks includes a customer response, a document summary and a work plan. Check the response for unapproved promises, the summary against the source, and the plan against the available time and dependencies. Participants should leave with work they have completed themselves, rather than having only watched a presentation.

Ask whether the duration means clock hours and whether it includes a break. My offer includes 230 minutes of work and a 10-minute break. See the 4-hour AI and ChatGPT training programme, with timings, exercises and expected outcomes for each part. It includes a Codex demonstration; hands-on use requires access agreed in advance.

Don't treat a short workshop as a promise to master all AI applications or implement ready-made automation. Select the most important applications and leave the more challenging tasks for further practice.

How do you compare two AI training offers?

Compare the same scope, rather than just the hourly rate. A training session with a presenter's demonstration and a workshop where everyone performs tasks may require different preparation and group organisation. Ask for a specific example of an exercise and a way to check its results.

For each offer, note what has been confirmed and what requires clarification. If the description doesn't specify materials, number of participants, or account costs, mark it as missing information. Don't assume that an unlisted item is included in the price.

Before choosing, send both providers the same description of your needs: participant roles and experience, two tasks to improve, and tool restrictions. This makes their proposals easier to compare. A suitable proposal should connect your needs to exercises and a clear way to assess the results.

• Objective: What task will the participant perform independently after the class?
• Practice: how much time is allocated for your own work and discussing errors?
• Group: how will the trainer support the planned number of participants?
• Tools: What accounts are required and who covers their costs?
• Materials: what will the participant keep and use after the training?
• Price: does it include preparation, travel and agreed materials?

Check whether participants will be able to practise

Before the workshop, confirm available computers, accounts, tools, and internet access. Determine whether paid features are required, who will provide access, and whether access will continue after the workshop. This will avoid situations where the method you've learned only works on the instructor's account.

It's worth completing a short practice task beforehand on the company network. Access restrictions, inability to upload a file, or login requirements can take up a significant portion of the session. The organizer should know who to contact to resolve such issues.

Agree which materials may be used for exercises and which tools may process them. If there are no established rules, use fictional examples. Training should encourage informed decisions about data, rather than indiscriminate document uploads.

Adapt the training format to the group

Ask how much time is allocated for independent practice, questions, and discussion of errors. A demonstration by a skilled person may seem straightforward. Only the participant's own work reveals what the participant doesn't yet understand.

The number of participants influences the ability to provide personalised assistance. There is no single group size appropriate for every programme. Ask for clarification on how questions will be collected and exercises reviewed based on the planned number of participants.

For on-site sessions, arrange room and workstation preparation. I conduct AI training at clients’ premises in Bydgoszcz and the Kuyavian-Pomeranian region. I confirm the date, programme and travel arrangements when we agree the scope.

How do you assess the results after training?

Choose a task the team will repeat after training. Before the session, record the current workflow and what a correct result looks like. Afterwards, use a similar example to compare the result and the time needed to check it.

Do not judge success only by the amount of text generated. Check whether participants can work independently, recognise errors and identify when they need another source. A short note of any difficulties helps guide further practice.

Ask what materials participants receive afterwards, such as prompt examples, instructions and assessment criteria. Recordings, consultations and follow-up support should be stated in the quote; they are not automatically included.

What should you share before the programme is prepared?

Include participant roles, experience, number of participants, and any tasks that could be improved in your inquiry. Indicate available tools and data limitations. For in-person classes, provide the location and organisational arrangements.

I use this information to suggest exercises that reflect your work. My service is described on the AI training for businesses and individuals page. If you later need custom software or an integration, that is a separate implementation with its own scope and price.