AI Design Sprints
AI design sprints that ship a decision
What changes when the teammate in the room is a model: faster prototypes, same kill criteria, no demo theatre.
The S-Curve·
// Skills
AI design sprints that leave a decision
An enterprise design sprint with a model in the room is still a sprint. Same kill criteria. No demo theatre.
An AI design sprint is a timeboxed sprint where the model is a teammate. Faster prototypes. Same kill criteria. If you leave with a demo and no decision, you ran a show, not a sprint.
People hear "AI" and forget the rest of the sentence. The rest of the sentence is design sprint. Problem, options, prototype, test, decide. The model does not get a vote. It gets a job: draft, generate, cluster, stub the first version so humans can argue about something real before lunch.
Enterprise design sprints already had this shape. The AI version is not a new religion. It is speed of options. I would rather see twenty ugly drafts in a morning than three polished slides on Friday.
How to run it
Frame the problem before anyone opens a tool. Then let the model flood the table. Then kill most of it. Humans test with humans. A facilitator keeps the room honest. Someone who owns the process has to be there. Someone who does the work has to be there. Someone who can kill the idea has to be there. A gallery of interested stakeholders will turn it back into a workshop.
Two days can be enough. Inkeros ran a two-day innovation sprint. Stretching to two weeks usually means nobody wanted to decide. The clock is a feature. Without it, the model will generate forever and you will call it exploration.
Design thinking is the parent method. Lean startup is the loop after the sprint if the idea deserves a longer life. Enterprise AI is the habitat: a hated workflow, not a playground. AI training is how the people in the room know what the model is for.
Infinitev generating landing pages to test ideas fast is the spirit: AI in service of a learning loop. Steal the loop.
What to train
You can teach the sprint mechanics and the AI craft in the same week if the problem is live. Collective Campus trains design thinking and sprints and AI skills for teams. Two links, because this hub sits on both. Do not buy a week of tool tourism and call it a sprint.
Geos: AU, NZ, HK, SG, the US, the UK and Canada. A sprint in a New York corporate is not a sprint in a Singapore bank. The clock is the same. The constraint is not.
What we publish here
AI design sprints that ship a decision. Design thinking that survives the board pack. Inkeros. Infinitev. The design thinking and AI training guides.
If your last "AI sprint" produced a demo and a parking lot, you did not sprint. You performed. Book a shorter room and a harder no.
The model will happily generate a strategy. Do not let it. Strategy is a human kill. Use the model for volume: sketches, copy variants, process maps, sample data, a first prototype. Then argue. The argument is the work. Teams that skip the argument because the output "looks done" ship plausible nonsense. Plausible is the danger. Ugly and testable is safer.
I would score a sprint on one question: what did we decide to stop, start or continue? If the answer is "we learned a lot," you learned how to avoid a decision. Write the decision on a page. Date it. That is the artefact. The prototype is evidence. The decision is the product.
Do not invite the whole leadership team to day one. Invite them to the decision. Leadership in the working room turns the sprint into a performance. The people who do the work go quiet. The model fills the silence with plausible slides. You will clap. You will not have tested anything.
What we have written
Design Thinking
Design thinking that survives the board pack
How to keep problem framing honest once risk, procurement, and a steering committee enter the room.
The S-Curve·
AI Training
CelcomDigi is asking Huawei to co-design how its people learn
A three-year memorandum in Petaling Jaya. Pathways in AI, data, 5G and cloud, and certifications at associate, professional and expert level. Albern Murty says advantage follows learning faster than the pace of change. Azmi Ujang, the chief human resources officer, was in the room.
The S-Curve·
AI Design Sprints
Singapore is measuring its AI training by what officers change, not courses finished
The Institute of Digital Government wants 150,000 public officers equipped by 2028, and it will judge itself on application and lasting workflow change. More than 1,000 officers have already been through its AI Sprints.
The S-Curve··
AI Training
Go1 is putting OpenAI Academy courses inside the corporate learning catalogue
The partnership puts OpenAI's own training in front of the 10,000 organisations already buying through Go1. Access gets easier. Deciding which work changes is still the organisation's job.
The S-Curve··
AI Training
Kier is judging its Oxford programme by what leaders take back
Fifty leaders. A three-day residential at Saïd Business School, then cross-business Growth Groups with Oxford. Stuart Togwell says humans stay at the helm, and the measure is how the learning is applied.
The S-Curve·
AI Training
HSBC is building an AI institute in Hong Kong to bring 200,000 staff along
The bank's chief executive says AI will transform some jobs, eliminate others and create new ones. Upskilling sits first on the new institute's list, with HKUST as anchor partner from early 2027.
The S-Curve··
Critical Thinking
Services Australia is growing its staff AI trial to 2,000 people, one thumbs up at a time
About 780 staff now ask the agency's how-to library questions in plain language, and their 28,000 ratings decide what changes next. The pace, the agency says, follows evidence rather than enthusiasm.
The S-Curve··
Field guides
Work in the room
2-day innovation sprint
Inkeros is a Canadian omnichannel sales and engagement company with 600+ employees across Canada and the United States.
The S-Curve · Jun 18
AI-powered landing page generator to test ideas fast
Infinitev is an Australia and New Zealand-based EV and hybrid battery specialist.
The S-Curve · Apr 11
Questions leaders actually ask
- What is an AI design sprint?
- A timeboxed sprint where the model is a teammate. Faster prototypes. Same kill criteria. If you leave with a demo and no decision, you ran a show, not a sprint.
- How is that different from a normal enterprise design sprint?
- Speed of options, not a new religion. You still define the problem, still test with humans, still decide. The AI does not get a vote. It gets a job.
- How long should it run?
- Two days can be enough if the problem is framed. Inkeros ran a two-day innovation sprint. Stretching to two weeks usually means nobody wanted to decide.
- Who needs to be in the room?
- Someone who owns the process, someone who does the work and someone who can kill the idea. A facilitator. A model. Not a gallery of interested stakeholders.