What happened at the AI-Design Sprint at the global digital festival, The Conference, two weeks ago:

I led the one-hour AI-Design Sprint with 21 participants, let me explain what we’ve done:

“Fashion” was chosen because participants were users as well, this allowed them to test each others solutions in the role of a user at the end.

This was the overview of the sprint:

The image below is the process of an AI-Design Sprint — it’s a funnel where you go from having many opportunities to use AI to your right AI solution for your business:

Phase 2, sketch & decide

Because our time was limited to one hour, usually the duration is a half day, participants started with phase 2 by being presented user needs&wants from phase 1 and in their role of user chose a need&want they felt is most important.

Then the participants made themselves familiar with relevant startups. The purpose is to understand where the fashion industry and its users are currently in order to rise above that in the sprint, and to detect trends.

The participants received it in this form:

The participants were then introduced to AI which in summary is and gives you super power:

Mapping out AI

AI was mapped out — about 100 (100!) individual specific AI functions on cards ordered in groups like “optimise complex systems”, “understand people using language”, “make predictions”, “create content”, “understanding what’s happening in pics and videos”, “autonomy”, “automation”, and “other technology” were placed on the ground.

All the mapped out AI technology is currently accessible, there isn’t a driverless car in it for example, so you can start implementing your solution today.

Here are a few of the about 100 AI cards:

Matching user need&want with AI

The participants then match their chosen need&want card with an individual specific AI technology card that could enable serving this need&want. Simple as that.

Practically, you have your need&want card in your one hand and pick one AI card and see whether that AI function could serve this need&want. If it doesn’t match, pick another AI card until you have a match.

To do so participants use a template where they fill in their chosen need&want on top, and on the left write the AI functions from the AI cards they think serve the above need&want. This might be one or several. Below you write in one sentence how your chosen AI functions could serve above need&want.

Building the solution around AI by starting with AI

The purpose of the sprint is not to add AI to a not well working old system, but using technology to question old ways of doing things and doing things in new ways for a best customer experience. Therefore participants use the hotel-reception-desk-analogy to build their solution around AI.

The hotel-reception-desk-analogy is that even though a hotel receptionists hardware got smaller, from desktop computer to laptop to tablet, the hotel reception desk strangely remains. Considering the possibilities of a tablet and removing obstacles between the receptionist and the guest, the receptionist should remove the desk and could check in the guest while the guest rests in a comfy chair sipping at his drink — having a great user experience.

Moving from a picture no. 3-solution to a picture no. 4-solution, participants write on the right side of their template what the “reception desk” in the area of their solution is.

They question their “reception desk”. Underneath participants write, starting with their chosen AI technology, how to use it for best customer experience in the area of their solution.

Phase 3, prototype & validate

The teams visualise their solution in form of a user story. The purpose is to create a first prototype to test the solution with real users.

Participants draw the user story of their solution in three pictures. It’s done by drawing simple stick figures, that simple, see below example.

Your first image depicts the user in a setting right before the use of your solution.

It might be helpful to see where in the overall user journey your solution takes place. Participants take the pre-user journey and mark with a pen where their solution takes place. For example, your solution could be in the “before” phase because your solution might be around personalised co-creation of fashion items.

The second image should depict the user while using your solution. The third image illustrates the immediate outcome for the user after using your solution.

Then teams test their solution on users by one presenter of each team presenting the teams solution to another team and seeking feedback. He finds out why users would use it or why not and what should be better. The presenter brings the feedback back into his team, they improve their solution by drawing a new user story.


The teams ended up with awesome solutions transforming fashion by applying AI, making fashion more human.

The participants used the last three minutes for individual reflection on their experience, answering, what happened?, how did it make me feel?, what will I do differently adding AI to a business tomorrow? This reflection time is where most of the learning happens, and some participants shared their insights with everyone afterwards.

The sprint was an amazing experience for everyone — without prior AI knowledge necessary they went from user needs to a user validated AI solution within one hour! That’s astonishing.

Thanks to all participants, you were awesome.


More info to AI-Design Sprints here.