A product squad can spend a quarter polishing a prototype, then tell the sponsor the quarter went well because the demo looked finished. The sponsor then funds a build nobody outside the room has tried. Before I book the days I ask what a customer did with the last version.
GV's sprint gives you a way to refuse that shape. You map the problem and pick a focus. You sketch competing options. You make the hard cut and write a hypothesis you can test. You build a prototype real enough to put in front of someone. You test it with customers before you spend the build. The week is a way to learn without launching. Friday can still kill the idea.
Where a model fits
An AI design sprint keeps that decision and puts a model in the slow parts. Research can get wider before anyone treats it as evidence. The room can see more options before it cuts. A thin prototype, a screen, a script or a walk through of the service, can exist before the days run out. Give the model the slow work. Keep a person on the check against something you can point at.
The AI design sprints workshop is this shape with a facilitator. The team leaves with a tested concept and a decision to proceed, to change course or to stop. I see the same failures often enough to name them before you book. A sprint that ignores the tools spends the week on sticky notes while a model could have sped the research and the first version. A demo with no decision lets someone show a chatbot while nobody decides what to build. A tool debate spends the days on which model to use, and never gets to whether the idea is any good.
The spine is still understand, focus, options, decide, prototype, test. The model compresses the slow middle. It leaves the test alone, and it leaves alone the person who may stop the work. A polished demo with no decision is a show. The sponsor who needed an answer still lacks one.
Who it suits
A product squad that needs a concept test, and that lacks six months of discovery, will get more from the days than a team chasing a certificate. So will an innovation programme whose sponsor has asked it to show progress on AI, and that is close to shipping a performance. So will the sponsor who has to fund the next step and who will sit in the decision, and who will skip the pack that arrives afterwards.
In Melbourne the room is specific. The AI design sprint in Melbourne is for the product or innovation team who will make the thing, and the sponsor who has to fund what happens next. Engineers help. You can run the days without them. A mixed team can still prototype. If you cannot name the customer or the operator who feels the problem, hold the booking. If you cannot name the person who may say stop, the days will produce a recommendation for a committee that missed the room.
Skip it if the team wants a certificate, a tour of logos or a vision for the year. One live problem is the brief. A portfolio of themes belongs in a different meeting, and it will spill past the days.
How the days are spent
We run the facilitated version in Melbourne. The core team meets in person. A sponsor can join from another city. The length is three to five days, or shorter when the calendar will refuse to move. The tools are the ones the team may use after the room empties. The brief is one live customer or operator problem. Leave this month's fashionable model off the brief.
You can run the classic five day arc, or compress it when you have locked the date. The jobs stay the same when the hours shrink. What changes is how strict you are about the cut. Before the first exercise, write the problem, the user and the decision the sprint must produce. Proceed, change course or stop. Include the evidence that would make you stop early.
Use a model to widen the research, then check what it produced against evidence you can point at. A fluent summary of a customer you have yet to speak to fails as research. Generate more options than the room would invent alone, and keep two. Twelve options is how a room refuses to choose.
The cut before you build
Prototype the thinnest version a stranger can react to. A screen, a script or a walk through of the service. Skip the new platform, and skip the vendor you will still be procuring next quarter. Put it in front of people who missed the room that made it. Keep it, change it or kill it, and record the decision where the next team can see it.
Data belongs in the brief. The session uses dummy or approved material, and it covers what must stay out of a tool. If nobody can say which records are in bounds, the sprint waits. Agree the rule with whoever owns it before the first prompt.
The test is there to catch the fluent mistake. A model will draft a confident version of the wrong service. The trial day is when that version meets a person who cares nothing for how tired the room is. If that slot becomes a demo to stakeholders, you have built a pitch.
What they leave with
A brief tied to a real customer or operator problem. A smaller set of options, because a model widened the set and a human threw most of it out. A prototype someone outside the room has reacted to. A decision to proceed, to change course or to stop. Guardrails for data and accuracy that the team used during the days, written down before the decision, and kept off the closing slide.
A stopped idea counts if you write the stop down. Organisations get stuck when every concept stays alive because nobody wants to be the person who ended it. Proceed means a named owner and a version the operator can run. Change means you know which assumption broke. Stop means the idea stays dead when someone tries to revive it next month under a new name.
If the sponsor needs another month of business casing before anyone may try the thing, you were preparing a meeting. Say that on the last day, while the people who did the work are still there.
What to book
This sits apart from the AI training day. Training builds skill across ordinary work. The sprint spends that skill on one live challenge and ends when someone decides to proceed, to change course or to stop. If the team has never used the tools, do the training first, or the sprint becomes a tour and the decision never arrives.
It also sits apart from a hack day. A hack day ends when the demo works. A sprint ends when someone decides. Afterwards you should be able to point at the decision and the evidence. The applause can fade.
A team that already has a decider, an operator and a data rule can run a thinner version inside the company. That sequence is in building an internal AI design sprint. Use it when you can hold the sequence yourselves. Use the Melbourne room when you want the sequence held and the decision made in the days you actually have.
What the test is for
To book, name the decision the sprint must produce, who can attend in Melbourne and any data you must keep out of a tool. Send that through the contact form. We shape the days around that constraint, and we leave our own case study at home.
A Melbourne team is ready for this when the problem is live, the sponsor can say stop, and the room will accept a killed idea as a result. If those are missing, book the meeting that creates them. The sprint will supply neither a decider, nor a customer, nor permission to kill the work.
The Google Ventures sprint is the five-day shape this week still follows.
