Industry
Creative Technology
Client
Spectrum
Role
Product Designer, Interaction Design
Spectrum - 3d Generative AI Application
Overview
Spectrum is a 3D creative suite with AI built into the workflow, designed for architectural visualization, gaming assets, and product prototyping. The goal was to help creators move from idea to execution without getting slowed down by technical complexity. Instead of making AI feel separate from the creative process, Spectrum was designed to feel like a natural part of it. The challenge was finding the balance between power and simplicity. Creative tools can become overwhelming fast, so the focus was on building something that felt advanced enough for professionals, but still approachable for newer users.
The Real Problem
3D and architectural visualization tools are powerful, but they often demand a lot of setup before real design work can begin. Users spend time dealing with lighting, materials, topology, and other technical details instead of staying focused on the creative part. AI can help, but it often feels disconnected or hard to control. I wanted to design a tool that felt more transparent, flexible, and useful in real-world workflows.
What I Designed
Spectrum uses AI to handle repetitive technical tasks while keeping the creator in control. The idea was to let AI support things like optimization, rendering, and material setup, while giving users room to make creative decisions themselves. That way, the tool supports the work instead of taking over it.
How It Works
Spectrum adapts to different levels of experience and different stages of the design process. A user can quickly generate a 3D environment from a rough floorplan, or manually refine details like lighting and materials when more control is needed. Everything stays non-destructive, so users can explore without locking themselves into one path. The experience was built around three ideas: Low entry point, high ceiling - Beginners can create quickly, while experienced users can refine details. AI as a partner, not magic - Users can see what the AI is doing, which helps build trust. Export ready - Outputs are production-ready and work across common pipelines like USDZ and GLB.
