After laying that groundwork, I found myself in new territory—less about training, more about experience.

When we created a car configurator in Unreal Engine, I had to learn how to build UI systems that responded fluidly in real time, all streamed from the cloud. This project opened my eyes to the potential of pixel streaming—how it can remove the hardware barrier and bring high-fidelity visualization to anyone with a browser. That was the moment I started thinking of real-time 3D not just as a development challenge, but as a bridge to accessibility.

The Chokola AR filters came shortly after. It was my first project with Spark AR Studio, and I had to understand how mobile AR works under the hood—how shaders behave, how tracking is managed, and how gesture recognition differs from body pose. Spark’s constraints forced me to learn optimization techniques I hadn’t considered before. And in doing so, I saw how even simple, playful experiences can make a brand feel personal.

Then came Creatures United. We needed to bring extinct animal avatars to life using only audio files that arrived a day before launch. That challenge pushed me to explore tools for audio-driven animation—creating automated facial and body motion from voice recordings. I didn’t just learn a pipeline; I witnessed how AI could save us hours of manual animation while still delivering expressive performances. That’s when it became clear: storytelling was no longer limited by timelines, only imagination.

In Power of Imagination, we aimed to shoot a film across multiple virtual locations—in just two days. I had to understand how to build real-time collaborative pipelines in Unreal Engine using Perforce and multi-user editing. I also learned how to synchronize nDisplay rendering with camera tracking, lens metadata, and physical hardware. What struck me most wasn’t just the complexity—it was how all these moving pieces could work in unison to support the creative process.

That lesson echoed in the music video we made with Komorebi. Her album’s comic book world became the canvas, and virtual production let us paint it into reality. Here, I learned how much power lies in blending the practical with the surreal. How a live-action shoot can live comfortably inside a CG dream—and how virtual production allows artists to iterate and visualize in ways traditional filmmaking never could.

And then there was Tsoma—a full-circle moment in a way. It was a small, self-initiated music video that used AI tools like ComfyUI and Stable Diffusion 1.5. I learned to sculpt visuals from prompts, to generate movement from audio, and to blur the lines between simulation and abstraction. In some scenes, I used Unreal to rough out geometry and animate particle systems in sync with the music. It was a poetic experiment—a tribute to how far generative tools had come and where they were going.