My role: AI Systems Designer
Scope: Algorithmic Pattern Design
Duration: January 2025- August 2025
Context
The fashion industry has always been governed by systems: grids, patterns, repetition, deviation. But these systems have been largely manual, relying on a designer’s hand and limited cycles of iteration.
With computational design and AI, those systems can be reprogrammed. Threads of Code reframes fashion as a computational process, treating code as a design material capable of inventing new silhouettes and forms that extend beyond traditional sketching or draping.
Challenge
Fashion today faces three challenges:
Creative repetition: Most collections repeat familiar silhouettes, recycling what’s safe and trending.
Slow, wasteful iteration: Each prototype requires physical sampling, consuming time, material, and labor.
Static identity: Clothing functions as a fixed object, not something adaptive.
Opportunity
The industry is at an inflection point. While digital tools are reshaping how products are sold and distributed, little attention has been given to rethinking the creative process itself. Computational design offers a chance to:
Accelerate creativity by generating thousands of possibilities beyond the limits of manual sketching.
Reduce waste by simulating textiles and garments before a single piece of fabric is cut.
Expand identity by enabling garments that reflect data, emotion, and context- moving clothing from static artifacts to adaptive systems.
Approach
First, I mapped how computational methods (generative algorithms, AI models, and parametric design) could intersect with traditional fashion workflows. Defined the project as a way to treat code as material rather than just a tool.
Algorithmic Pattern-Making
I designed generative systems using Voronoi grids, L-systems, and noise functions to produce unpredictable yet structured textile patterns. By harnessing “controlled randomness” to simulate the organic irregularities of nature, I turned algorithmic variance into unique prints.
AI-Driven Silhouette Generation
I trained AI models on base garment structures to produce unconventional draping, layering, and form.
Used iterative curation to filter out the noise and surface silhouettes that balanced provocation with wearability.
3. Bridging Digital and Physical
I translated selected AI outputs into small-batch textile samples and 3D garment prototypes. I tested how computational aesthetics lived in fabric, iterating between screen-based simulation and tactile reality.
Output
Digital Lookbook: A curated collection of algorithmically generated prints and silhouettes, presented in 3D.
Fabric Swatches: Small-batch textiles produced from computationally designed patterns. These swatches tested how algorithmic aesthetics translated off-screen, allowing for material, texture, and color validation.
Conceptual Garments: Prototype pieces that embody the project’s vision- garments shaped by algorithmic processes yet curated through human direction. They function as proof-of-concept for how computation can move from simulation into real-world wearability.
Key features
Code as Material
Algorithms weren’t just tools for rendering. They acted as the raw material, shaping textiles and silhouettes from the ground up.
Controlled Randomness
Designed systems where unpredictability became a creative driver, producing organic irregularities that mimic nature rather than sterile precision.
AI-Human Curation Loop
The process positioned AI as a generative partner. Hundreds of outputs were produced, but I curated which ones advanced to prototype.
Digital-to-Physical Translation
Moved beyond screens by testing computational outputs on fabric swatches and garment prototypes, proving that algorithmic aesthetics can survive in real-world textiles.
Sustainability Through Simulation
Reduced reliance on repeated fabric sampling by iterating virtually, saving time, cost, and material waste.
Behind the build
This project started as a question: what happens when you treat code like fabric? I approached it less like a traditional fashion collection and more like building a system that could surprise me.
Experimentation over certainty
I wrote algorithms knowing they would misbehave. Some patterns looked broken, chaotic, even unusable, but those “errors” became the starting point for prints that felt right.
Curation as design
AI produced hundreds of silhouettes, most unwearable. My role became less about generating and more about deciding, choosing which outputs carried enough tension between provocation and possibility.
Iteration loops
I bounced constantly between screen and material. A pattern that looked compelling in code sometimes fell flat on fabric; other times, a texture I almost discarded came to life when printed.
Impact
Threads of Code reframed fashion as a system rather than a static craft. By merging algorithms with design intuition, the project:
Generated forms and patterns that a traditional sketch-first process could never reach, proving computation can open radically new design directions.
Demonstrated how virtual iteration can cut down on unnecessary sampling, pointing toward a more sustainable model for early-stage fashion exploration.
Framed clothing not just as an artifact but as the output of dynamic systems: adaptable, data-driven, and responsive.
Reflection
This project reminded me that design is as much about letting go as it is about control. Working with algorithms meant embracing unpredictability, allowing systems to surprise me rather than forcing outcomes. The role of the designer shifts here, from maker to curator, from controlling every line to shaping the dialogue between human intention and machine output.
This project also sharpened my perspective on where fashion can go next. It’s not just about garments that look different; it’s about processes that waste less and reflect identity in new ways. For me, this was a statement about how design itself is evolving: toward systems, toward hybridity, toward futures we haven’t yet fully imagined.