SideKick
SideKick began as a personal exploration into generative image AI and React development. What started as a learning initiative quickly evolved into a deeper investigation into how AI can complement, rather than disrupt, the creative workflows of 3D designers. By bridging the gap between AI capabilities and existing design processes, SideKick empowers industrial designers, 3D artists, architects, and professionals across other 3D digital mediums to seamlessly incorporate generative technology into their workflows.
2024
Year:
Sole Designer & Engineer
Role:
Prototype
Base Concept
SideKick is built to support the messy, exploratory nature of early-stage design. Instead of relying on prompt-based generative AI, it adapts to how designers naturally work, allowing them to experiment and iterate freely.
The focus is not on producing polished final images, but on helping designers express their ideas and intent throughout the creative process.
Feature Highlights
Real-time collaboration:
Generate images as user’s work is refined and developed.
No plug-in required:
Through browser’s shared screen, users can utilize all their existing creative tools.
No additional preparation:
By using ControlNet the AI can interpret to any condition of model/image.
Path Forward…
The next steps for SideKick focus on enhancing the flexibility and usability of the tool.
Test various baseline models and fine-tuned models to reduce reliance on a single third-party API, enabling greater customization and control for the user.
Explore the limitations of using pixel-based input to generate images, particularly with screen capture interactions, which may struggle to accurately interpret and generate complex 3D output.
Fine-tune user input and control, providing users with essential options that fit their workflow without overwhelming them with excessive settings, streamlining the design process.
Tools:
React, Tailwind, Figma, Replicate, Comfyui, Rhino3D & Maya