Private vs public AI training: a seat, a team session, or having it built

Buy one public seat when one person owns the task. Book private onsite training when several people do the same job and the output has to look the same whoever writes it. Buy neither when nobody on your team wants to learn the workflow, because then you want the system built for you, and that is a separate engagement. The arithmetic that settles most of this is plain: A$495 excluding GST per seat, multiplied by the number of people who would genuinely be working for the whole a full day.

TL;DR

• Count the people who would be working, not the people who would be in the room. Private training for a job one person already owns pays for an audience.
• Public seats are linear. Four people is A$1,980 including GST, eight is A$3,960. Private training is a fixed quote after scoping, so the seat total is the number you hold the quote against.
• One seat is the right first purchase while you still don't know whether AI helps your work at all.
• A private session earns its price when a group has to agree one way of doing a repeated job, because the agreement is the thing you keep.
• The public workshop takes a maximum of 15 paid attendees, and those seats are open to anyone who books, so a team of 12 can't assume it will get the room.
• If nobody wants to own the workflow afterwards, neither training format is the right purchase. We build and connect systems too, as a separate engagement with its own scope and price.

What actually differs between a public seat and a private session?

Three things: the room, how the price is set, and who decides what gets worked on.

The teaching is the same in both rooms. You write the five inputs that make up a reusable prompt, which are Context, Task, Source, Rules and Output, then test them on a real piece of work and check the draft against the material you supplied. You leave with the instructions and the example you tested.

What changes is the ownership of the agenda. A public seat is a full day on your task in a room where other people are working on theirs. A private session is built around what your team does, which means the scoping conversation decides the content before anyone turns up.

Do the seat arithmetic before you enquire

Public seats are one of the few prices in this market you can work out on paper, so work it out.

Two things that column doesn't include. The first is time. A full day per person, plus travel each way, at whatever an hour of that person's day is worth to you. For a team of six that is most of a working day gone before you count the seats, so put your own hourly figure against it rather than guessing. The second is the cap. The public workshop takes a maximum of 15 paid attendees and the seats go to whoever books, so a business sending eight people is hoping eight seats are still open on a date that hasn't been announced yet.

I can't tell you the exact number where a quote beats the seats, because the private number comes out of scoping rather than a rate card. Team size, session length and travel set it. What I can tell you is which number to hold it against: run the seat sum for the people who would actually be building something, add their time, then compare it with what AI training costs across the market and ask us for the quote. If the quote comes back above the seat total and all you need is one or two people building a workflow, take the seats. If it comes back above and what you need is a room full of people agreeing one way of working, the difference is what the agreement costs, and that is a decision about value rather than price.

When is one public seat the right buy?

When the work lives with one person, or when you don't yet know whether any of this helps.

Four situations point at a seat rather than a session.

The first is a task that already lives with one person. Quoting sits with the estimator. Enquiries sit with whoever answers the phone. Our work with Glen House Homoeopathy covered SEO, enquiries and admin, and in a practice that size those jobs belong to one or two people. Training eight people to build a workflow for a job one of them does is money spent on an audience.

The second is testing the idea at all. A$495 and a full day is a cheap way to find out whether AI is any good at the thing you keep doing by hand. Send the person most likely to use it, not the most senior person in the business. If it works, you'll know what to ask for next and you'll have someone internally who can describe it.

The third is a team whose people do genuinely different jobs. A team session is strongest when everyone is working on the same repeated task. Where the bookkeeper, the estimator and the receptionist each have their own problem, tell us at scoping so the session is planned for that, or send one or two of them to the full-day public workshop and let them bring back a method the others can copy.

The fourth is a business where only one or two people will ever do the building. Plenty of them have one person who likes this sort of thing and a team who would rather be handed the finished instructions. Buy the seat for the builder.

When does a private session earn its price?

When several people have to end up doing the same job the same way, and that agreement has to happen with all of them in the room.

The clearest case is a task that moves between roles. We helped automate Radiant Roof Repairs' internal processes with AI, and internal process work is exactly this shape: information starts with one person, gets handed to another, and comes back out as a document a customer sees. Nobody can fix that alone. The person who writes the notes has to agree what goes in them, because the person drafting from those notes is stuck with whatever they get.

The second case is consistency. If five people each send customer replies and all five read differently, the point of the session is one agreed set of rules and one agreed source of facts. You can't delegate that to one attendee and a handout.

The third is your own material. Private onsite training runs at your workplace, on the laptops and accounts your team already uses, built around your tasks and the output you need. A fixed quote follows a conversation about the team's tasks, location and experience, and tool and account requirements are settled in that same conversation.

The fourth is timing. A private session gets scheduled through scoping. A public seat waits for a date that hasn't been announced, which matters if you need people trained before a busy season rather than whenever the next workshop lands.

Where can each format actually happen?

Private onsite training happens where your team is. The public workshop happens at Mooloolaba.

We run private onsite AI training on the Sunshine Coast and in Brisbane, Gold Coast, Ipswich, Noosa, Moreton Bay, Logan, Redlands, Toowoomba and Gympie. On the Sunshine Coast that covers Maroochydore, Caloundra, Mooloolaba, Buderim, Birtinya, Nambour and Coolum, so a team based in Buderim gets the session at its own desks rather than in a hotel function room.

Distance changes the sum. One person from a business in Logan can drive up for the public workshop and it costs a seat plus a day. Send six people on that drive and it costs six seats plus six days, which is the point where travel becomes a real argument for having us come to you. Travel is one of the three things that set the private quote, so it lands on the invoice either way. The question is whose travel it is.

What can you safely work on in each room?

A safe example in the public room. Your own material in your own building, under rules your team has agreed.

As a matter of best practice, the Office of the Australian Information Commissioner recommends that organisations don't enter personal information into publicly available generative AI tools, and it puts particular weight on sensitive information, because of the privacy risks involved 1. The recommendation applies in your own boardroom exactly as it applies at Mooloolaba, and it is why the public workshop asks you to bring a safe example alongside the laptop and charger.

What a private session adds is the decision itself. With the team in one place you can settle which categories of information go into a tool and which stay out. You can agree what the reviewer checks before anything reaches a customer, then write the whole lot down as a team AI use policy. The OAIC asks organisations to establish policies and procedures for the use of AI systems as part of good privacy governance 2. A rule the team wrote while building the workflow survives longer than one emailed to them afterwards.

The checking step is identical in both rooms, because these tools produce inaccurate or false results and don't understand the material they handle 3. Compare names, dates, figures and claims against your source. Recalculate totals. The person responsible for the work approves the final version.

Training is the wrong purchase when nobody wants to own the workflow

Ask who is going to run and maintain this in six months. If the honest answer is nobody, then training is the wrong thing to buy.

Training and implementation are two rungs of the same ladder here. The public workshop and private onsite training teach your people to build and run a workflow themselves. Implementation is where we build and connect the system for you. Both are ours and both start with the same enquiry, so the job before you send it is working out which rung you actually want.

Ownership decides it more often than complexity does. A business with someone who enjoys this work and has time for it gets more from training, because the person who built the workflow is the person who can fix it when a supplier changes a form or a price list moves. A business where every capable person is already flat out will buy training, use the workflow for three weeks and quietly stop, and then the cheaper purchase turns out to have been the expensive one. Be blunt with yourself about which one your business is, and treat it as a different question from choosing between training and paying for advice.

There is also a sensible order. Take one seat, build one workflow, find out whether the idea holds up on your own work. If it does, you'll know exactly what you want built, and describing it accurately is most of what makes a scoping conversation quick. Implementation is a separate engagement with its own scope and price, so tell us what you want built when you enquire and we'll scope it from there.

Frequently asked questions