One designer and hundreds of assets (Category #AI)

What caught my attention in “One Designer + AI. Hundreds of Deliverables” was how the designer at the company that runs the AI Engineer conference responds to a large workload. He doesn’t turn up his nose with an “ugh, this isn’t like Figma” (that’s my interpretation of a possible attitude toward all these AI tools:)), but changes his processes and builds tools that help him handle the volume. Vincent Wendy, a designer at AI Engineer, describes how he prepares conference materials. According to him, they have one designer for 7000 attendees and more than 300 speakers. Cards, schedules, wayfinding—all of it needs to be assembled, checked, and updated on time.

He starts with rules: fonts, colors, components. Then, with Devin, he builds a speaker-card generator and a tool for assembling schedules from current data. Repetitive work becomes a matter of selecting the right data and exporting the finished image. I like the thinking behind this. When you keep assembling similar materials, you can put some of that effort into a tool that will help you do it again. The designer knows exactly where his time goes and which details are easy to miss. That knowledge becomes the basis for a task to give the agent.

There's also a good moment in the talk: a schedule needs an urgent correction, but the tool has no edit button. Wendy asks Devin to add one, changes the data, and exports a new PNG. The process keeps changing as real working situations arise.

That willingness to rebuild the way you work is what interests me. The familiar tool remains useful, while the range of possible solutions grows. And building a small application for your own task becomes part of the designer's work.

P.S. I mentioned this case in the 3 AImigo episode on the AI-native organization.

#AI #Agents #Engineering #Product #Conference