Services
AI Workshops
Half a day. A handful of companies in one room. One seat at the front. You bring the process that hurts, we diagnose it in front of everybody, and then we spend most of the day building the fix with your people.
The room
One company in the seat. Everybody else in the room, taking notes on a problem shaped like theirs. Then the seat moves. Company types shown for illustration: attendees are never named without their say-so.
Booking runs through our standard scheduling form. It asks which process is costing you the most, which is the same question the hot seat opens with, so it is worth answering properly.
What is a Code and Trust AI workshop?
A Code and Trust AI workshop is a half day session where several companies share one room and take turns in a hot seat. Roughly 20% of the time is teaching, 20% is diagnosis, and 60% is hands-on building. Each company leaves with a working automation started in its own hot seat, a prioritized AI roadmap, prompts and playbooks, and the session recording.
There is no fixed curriculum, because we have never met a company whose bottleneck matched the deck. We work like a special ops team: hear you out first, then give you exactly what your situation needs. Sometimes that is a two hour build. Sometimes it is twenty minutes of being told the thing you wanted to automate should be deleted instead.
The format is borrowed from the way the best operator workshops run, and it is the opposite of a training course. Nobody presents at you for a half day. The room works.
How does the hot seat format work?
In the hot seat format, one company at a time describes the process that is costing it the most, gets questioned until the real bottleneck is identified, and then builds a fix live with a Code and Trust engineer while the other companies watch and map the problem onto their own operations. Then the seat moves to the next company.
You take the seat
You describe the process, not the vision. Where it starts, who touches it, where it stalls, what happens when somebody is on vacation. We ask questions until the bottleneck on the board is the real one, which is frequently not the one you walked in naming.
The room works while you are up there
Nobody is politely spectating. The other companies are mapping your problem onto theirs, and they interrupt. Most attendees tell us the seat they learned the most from was somebody else’s, because it is far easier to see a process clearly when it is not yours.
Then we build it
Your operators and a Code and Trust engineer build against that bottleneck, live, with the room watching. This is the part that consumes most of the day, and it is the part almost no AI workshop actually does.
The seat moves
The next company goes up. By the end you have watched several diagnoses and several builds, only one of which was yours, and the pattern between them is usually the most valuable thing anybody carries out of the room.
This is the reason we do not run it privately by default. A room of one company only ever finds that company’s blind spots. A room of several finds each other’s.
How is the half day actually spent?
A Code and Trust AI workshop is roughly 20% teaching, 20% diagnosis, and 60% hands-on building. Most AI training inverts that ratio and spends the day explaining what AI could theoretically do. Here the explaining is the small part, and the majority of the room’s time is spent building against a problem that was named out loud an hour earlier.
- 20%Teaching
- The smallest slice on purpose. Enough to level-set what the current tools genuinely do, and more usefully, where they quietly fail.
- 20%Diagnosis
- Each company in the seat, describing the process that hurts, until the real bottleneck is on the board instead of the one everybody assumed.
- 60%Building
- The majority of the room’s time, hands on keyboards, building against the bottleneck that was just diagnosed.
What do you leave with?
Every company at a Code and Trust AI workshop leaves with four things: a working automation started during its own hot seat, a prioritized AI roadmap specific to its business, the prompts and templates and playbooks used on its problem, and the full session recording plus materials. All four are yours whether or not you ever hire us for anything else.
A working automation
Built during your turn in the seat, running against your real process. Half a day with several companies in the room does not produce a hardened production system, and we are not going to tell you otherwise. It produces a real working thing, proof the approach fits your process, and the person from your team who helped build it.
A prioritized AI roadmap
Specific to your business, ranked by what the process costs you today rather than by what is interesting to build. It is the document you take to the people who were not in the room.
Prompts, templates, and playbooks
The exact prompts we ran at your problem, plus the templates and playbooks behind them, written down and yours to keep. Reusable by people who never attended.
The recording and every material
The full session, including the other companies’ hot seats. Your operators will rewatch the build. Your leadership will rewatch the diagnosis. Both are worth having.
Why not just use ChatGPT?
Everybody in the room already has a subscription, so the subscription was never the constraint. What a facilitated AI workshop adds is knowing which process is worth automating, knowing where the model quietly fails, getting past the first integration into your real systems, and putting the right people in a room with uninterrupted time. None of those come with a seat license.
Knowing which process is worth automating
Teams automate the annoying process rather than the expensive one. They are rarely the same process, and the difference is the entire return.
Knowing where the tool quietly fails
Output that looks right and is wrong is the failure mode that costs real money. Knowing where a model stops being trustworthy is not something the chat window volunteers.
Getting past the first integration
The chat window is where AI demos. Your systems are where it has to work. Nearly every stalled AI effort we see stalled at exactly that seam.
A forcing function
Most AI projects do not die at the idea. They die because the right people never got uninterrupted time in the same room. Booking the half day is the intervention.
Put it another way: the gap between a company getting value from AI and a company talking about AI is almost never the model. It is the hundred small decisions between the idea and something that runs on a Tuesday without anybody babysitting it. That is what the seat is for.
How is this different from every other AI consultant?
Most AI consulting sells strategy: an assessment, a maturity model, a roadmap deck, and an invoice. A Code and Trust AI workshop sells the opposite half. We build during the session, in front of you, with your operators on the keyboard, because the part companies actually get stuck on is not deciding to use AI. It is the first working thing.
Most AI workshops teach the tools.
We start from your process. The tools are ten minutes. Your process is the whole problem, and it is the thing nobody else is willing to sit in.
Most engagements end with a deck.
This one ends with something running. A deck is a promise that work will happen later. We would rather do a smaller amount of the actual work in front of you.
Most sessions are sold to leadership.
We insist the operators come too. The moment worth paying for is the one where the person who owns the budget watches the person who owns the process build the fix.
Most consultants want a long engagement.
We want you to leave able to do the next one without us. If you still need us after that, it will be for something genuinely harder, and you will know exactly what you are buying.
What tends to end up in the seat
We have run these workshops. We are not going to dress that up with a client logo wall we have not cleared or numbers we cannot show you the math on. What we can tell you is the shape of the problems companies bring, because they rhyme.
- A regional manufacturerQuoting. The quote itself is rarely the bottleneck. The bottleneck is the six places the inputs live and the one person who knows how to reconcile them.
- A national logistics firmException handling. The happy path was automated years ago. Everything expensive now lives in the cases that fall out of it and land in somebody’s inbox.
- A multi-office professional services groupEverything that happens between winning the work and starting the work. Intake, kickoff, and the same document assembled by hand for the hundredth time.
Who should you bring?
Bring leadership and the operators who will actually use the tooling, in the same room at the same time. Leadership alone leaves with a plan nobody executes. Operators alone leave with a tool nobody funds. The half day is built around having both present while the fix gets built, which is why attendance is a qualification question rather than a formality.
You do not need engineers. The building is done with tools your operators can keep using on Monday, facilitated by a Code and Trust engineer. If you have technical people, bring them and you will get further in the seat. If you do not, the session was designed for you anyway.
Read this part first
Who this is not for
This is not a workshop for companies that are curious about AI. It is a workshop for companies that can name a process that hurts.
If you cannot finish the sentence “the thing costing us the most time every week is _____” with something specific, the hot seat has nothing to work on. You would spend a half day watching other companies get value, and we would rather not sell you that.
- Teams that want a general AI literacy session for the whole company. This is a working session for a specific problem, not an all-hands.
- Anyone who needs a vendor evaluation, a tooling comparison, or a procurement document. Useful work. Different work.
- Companies where nobody attending can approve a change to how the work gets done. The build lands and then nothing happens to it.
None of that is a criticism, it is a sequencing problem. Come back when a process is visibly bleeding, or start with an AI audit, which is built for exactly that stage and will tell you which process to bring.
What does it cost?
Code and Trust quotes AI workshop pricing on the call, once we know who you would bring and what problem would go in the seat. The half day is fixed. What it is worth is not, because that is set entirely by the problem you bring, and publishing a single number would make it wrong for almost everybody reading this page.
That is not evasion, and you will have the number before you commit to anything. When you hear it, hold it next to two things:
What the process costs you now
Most teams have never actually added it up. Hours per week, times people, times fifty weeks, plus the work that quietly does not get done because those hours went somewhere else. Do that arithmetic before the call. It usually settles the question on its own.
What the alternative path costs
A full AI implementation engagement is a much larger commitment of money, calendar, and internal attention. The workshop is the cheapest way to find out whether your problem is worth that, before anybody signs up for it.
AI workshop FAQ
Common questions about Code and Trust AI workshops cover the format (half a day, several companies sharing a room, hot seat by hot seat), who to bring (leadership plus the operators), what you walk out with, and how confidential material is handled in a shared room. Pricing is quoted on the call rather than published.
How long is an AI workshop?
Half a day. That is deliberate. A full day of AI training is mostly a full day of people checking Slack, and a ninety minute session never gets past the diagnosis. Half a day is long enough to build something and short enough that the operators you need in the room can actually be released for it.
Is the workshop private to my company?
No. Several companies share the room and take turns in the hot seat. That is the format, not a cost-saving compromise. A room of one company only ever surfaces that company’s blind spots. A room of several surfaces each other’s, and most attendees tell us the seat they learned the most from was not their own. If you need a closed-door engagement, that is a different conversation and we are happy to have it.
Who should we send?
Leadership plus the operators who will actually use the tooling. Both, in the same room, at the same time. Leadership alone leaves with a plan nobody executes. Operators alone leave with a tool nobody funds. The half day is designed around having both in the room while the fix gets built.
Do we need engineers or technical people to attend?
No. The building is done with tools your operators can keep using on Monday, and a Code and Trust engineer facilitates. If you have technical people, bring them and they will get further. If you do not, the session is still built for you.
Will we leave with a system running in production?
You will leave with a working automation started during your hot seat, running against your real process, plus the person from your team who helped build it. We are not going to claim a half day with several companies in the room produces a hardened production system. What it produces is a real working thing, proof the approach fits your process, and a clear path to finishing it.
What about our confidential data?
You control what goes on the screen. We work from your real process, and in a shared room you decide the level of detail you are comfortable putting in front of the other companies. Anything you would rather not show them, we work through with you separately after the session.
How much does an AI workshop cost?
We quote it on the call, after we know who you would bring and what problem you would put in the seat. The half day is fixed. The value is not, because it is determined entirely by the problem you bring, and publishing one number would make it wrong for almost everyone who reads it.
What if we are not sure we have a problem worth bringing?
Then this is not the right session yet, and we would rather tell you now. The workshop needs a named process that hurts. If you are still at the stage of figuring out where AI fits at all, start with an AI audit instead, which is built for exactly that.
Related services
The workshop is the fastest way in. When a hot seat turns up something bigger than a half day can finish, these are the paths it usually leads to.
What would you put in the seat?
That is the whole call. If you can name it, we can tell you in about fifteen minutes whether a half day fixes it, and if it does not, we will tell you that too.