Pixel Portraits: Blending AI and Playful Nostalgia
The Activation
At our recent Next.js Conference and ShipAI events, we introduced an activation that blended technical experimentation with playful nostalgia. The idea started long before anyone stepped into the venue. As part of the online registration experience for both events, attendees could prompt and generate their own trading cards, giving them an early taste of the format and creating the foundation for what we wanted to bring into the real world.
A Pixel Portrait Trading Card System
Experienced designers crafted a pixel portrait trading card that felt personal while scaling to thousands of attendees. At its core was an AI-powered photobooth capturing each participant's portrait and transforming it into a custom pixel-style rendering, then printing it on demand as a collectible card. Every card was sealed in a PSA-inspired sleeve and finished with a serialized grading label unique to each attendee.
Behind the Simplicity: A LightWEIGHT System for User-Driven Storytelling
What emerged was more than just a keepsake. It became a lightweight system for generating user-led storytelling, where attendees stepped into the brand and carried it forward themselves. As the cards spread across social channels, they turned into small but striking symbols of the events, each one a pocket-sized artifact of the experience we wanted to create technologically sharp and unmistakably ours.
The Technical Pipeline: Invisible Effort, Visible Impact
Behind the simplicity of stepping up to a photobooth and receiving a trading card was a tightly orchestrated pipeline spanning AI generation, image processing, and on-site production. The flow began with someone posing for their photo. The captured image was sent into our generation pipeline, which used the flux context model paired with custom LoRA training on an nostalgic 16-bit pixel art style. The system produced a pixel portrait that preserved the attendee's likeness and rendered it onto a bright green background for downstream processing.
From Image to Card: The Assembly Line of Creativity
From there, BRI's RMBG-20 model removed the green background, isolating the character for compositing. The Canvas API assembled the trading card by layering the pixel portrait with the event's background pattern and brand overlay. A custom gradient map applied with the sharp image library created the final color identity for each event. The finished PNG was then sent to the on-site printer, completing the digital workflow in roughly six seconds.
Scalability and Attendee Engagement
Attendees could also scan a QR code to save a digital version of their card. The physical side of the workflow completed the experience. Each printed card was hand and die-cut into a rounded corner profile familiar from classic trading cards.
From Event to V0: A Browser-Powered Experience
After the events, we wanted Pixel Portraits to live as more than an in-person moment. We built a V0 template that brings the same AI-driven experience into the browser powered by the Vercel AIGateway and the Fal integration. The template allows users to take a photo with their webcam or upload an image. The app crops it to a card-friendly aspect ratio and converts it into a base64 data URL style reference, which serves as the visual anchor for the desired 16-bit look GPT-15 Edit model.
Remixing the V0 Template: From Event to Festive Variants
The template became a natural foundation for seasonal experiments. For the holidays, we created Vercelf Yourself as a spin-off that turns your photo into a festive 16-bit pixel portrait. Vercelf Yourself uses the same style reference approach but extends it with multiple filters. Each filter—like Vercelf Comfy Santa Wham—has its own reference image and tailored prompt aiming for a strict grayscale aesthetic.
The Core of the Request: Fal Integration
The app uses Gemini image generation through the Vercel AIGateway and the AI SDK. This setup allows us to experiment with models or update the stack later without rewriting the app, keeping the integration minimal. That leaves us free to focus on what the experience feels like, not how it's wired together. Create your own festive portrait in the Vercelf Yourself app—you can generate up to three images per day or remix the V0 template to experiment with filters, prompts, and models using your own AIGateway setup.