PocketPro
A personal AI co-pilot that helps you create, iterate, and play with music.
*Please note that all images and videos are AI-generated for this project
Context
Google Labs Creative Makeathon
Role
Product Designer
Team
1 PM, 2 Engineers
Time
24 hours
Context
What if music creation was as spontaneous as inspiration itself?
Creative ideas rarely happen at a desk. They emerge while walking, listening, observing, or simply existing in the world. Yet most music production tools require time, setup, and technical commitment—causing many spontaneous ideas to fade before they’re ever captured.
What if music creation was as spontaneous as inspiration itself?
PocketPro is a pocket-sized AI co-producer designed for those moments.
By combining AI sound generation with conversational interaction, PocketPro transforms everyday inspiration into playable music on the spot.
The Problem Space
Creativity isn’t stationary.
Music tools are.
Music creation tools today fall into two extremes:
Professional digital audio workstations (DAWs): offer depth and control but require time, setup, and technical commitment
Mobile music apps: accessible, but often limit deeper iteration
Meanwhile, generative AI is rapidly entering music creation, with the market projected to reach nearly $2.8 billion by 2030—yet most tools still treat AI as an output engine rather than an active creative partner.
Why music co-piolit?
PocketPro reframes AI as a music co-pilot. Inspired by early projection keyboard technologies, the project imagines a lightweight, portable interface that turns any surface into a playable space.
By embedding AI directly into the act of making music, PocketPro proposes a more continuous, playful, and human-centered creative experience.
Design Process
Designing with AI as a creative partner
Generative AI tools allowed us to rapidly explore form, interaction, and identity—helping us visualize what a future music co-producer could look and feel like before committing to a single direction.
Visualizing the Hardware
Using Google Lab's ImageFX, we explored different forms, scales, and material qualities to imagine a music tool that felt both portable and expressive.
These explorations helped ground the speculative concept in a tangible form.
When AI Image Generation Gets Hard
While AI excelled at abstract and atmospheric visuals, it often struggled with consistency, proportions, and functional detail when translating abstract prompts into precise hardware designs.
These failures became useful signals, revealing where AI lacked structural understanding and where human design judgment was essential.
AI misread the keyboard as a DJ setup, shifting the focus to a crowd.
Hardware appeared, but in the wrong form—and the scene unexpectedly moved to a subway.
The crowd disappeared, but AI loves the subway setting.
With a clean background, the AI defaulted to a traditional piano keyboard layout.
Runway Video Generation Experiments
We also experimented with AI-generated video to explore how PocketPro might be used. Although many clips failed to clearly communicate interaction or motion, they helped surface gaps in the experience and informed where stronger design intent and motion prototyping were needed.
AI introduced an unexplained string connecting the user’s finger to the device.
A remote control magically appears and disappears.
User Interface Design
With insights from these explorations, we refined PocketPro’s interface to be minimal, ambient, and rhythm-driven.
The UI supports quick sound layering, looping, and iteration, allowing users to engage with music playfully without the friction of traditional production tools.
Final Result
PocketPro — Beats in your back pocket
PocketPro reframes music creation as something fluid, accessible, and embedded in everyday life, lowering barriers and expanding who gets to create.
I produced this video to demonstrate PocketPro’s core interactions and communicate its creative vision.
Speculative Track Winner:
Greatest Potential to Transform
PocketPro was awarded for the potential to redefine how people interact with
sound, creativity, and AI!
A special shout-out to my teammates: Elvira Lu, Nick Marker, and Allison Huang.
We did it!
















