AI Videobooth
Walk up to a booth, take a photo, and walk away with a printed AI-enhanced portrait and a personalised AI-generated video. That’s what over 500 people did at Google Cloud’s AI Nordics Tour.
The Product
We designed and built the entire system from scratch for a startup we’ve worked with for the past year. The flow is simple: an attendee steps up, takes a photo, and the system kicks off an AI pipeline — the image gets enhanced, a physical print comes out instantly, and in the background a personalised video is generated and sent to their phone. The whole thing takes minutes.
Its first real-world deployment was Google Cloud’s AI Nordics Tour — a four-country corporate series across Denmark, Sweden, Norway, and Finland. Over 500 attendees used it, 800+ photos were printed, and the booth ran stable throughout long event days with high foot traffic. Many attendees described it as one of the most memorable parts of the tour.
How It Works
The booth interface runs on Vue.js and Electron. Photos get uploaded to Firebase, enhanced through a custom AI workflow on Modal.com, and videos are generated using Veo3. We added image analytics and sentiment analysis via Google Gemini to give each creation a personal touch. The entire backend runs on Firebase Functions and Google Cloud — built to handle concurrency, offline fallbacks, and the kind of reliability you need when a Google Cloud logo is on the booth.