The future of fashion:
Designing an AI autonomous
trend-to-product engine
Client: Confidential Global Fashion Retailer
My Role: Product Designer
Duration: 3 months
A large global fashion retailer approached us with an opportunity to reimagine how fashion and apparel are designed, bought, and made in the autonomous age. The retailer recognized several key needs: staying ahead of cultural trends, capturing a significant share of the lucrative Gen Z market, increasing speed to market, and developing a leapfrog model to outcompete fast-fashion giants like SHEIN while reducing waste.Given the rapid adoption of Generative AI across industries, our client saw a strategic opportunity. By leveraging AI for its efficiencies, they aimed to not only stay competitive but also to innovate in ways that only a business operating at a cross-category portfolio level can. This approach would enable them to capitalize on their unique position as a global retailer and lead the market into the future.
The challenge
The client wanted to explore the opportunity of building a working prototype of an Autonomous Innovation Engine together.
Strategy and my role
My team and I at BOI developed an AI-powered engine that prioritized on-trend, feasible garment production over volume, using real-time data, AI agents, and custom Gen Z synthetic personas to ensure new concepts met time, resource, and brand identity criteria while prioritizing collections by purchase intent. As a product designer, my role was to design a user-friendly UI and UX that condensed the real-world design process from six months into a few clicks.
When designing the engine interface, my first priority was to gain a thorough understanding of the real fashion design process and the end users on the fashion team. This involved designing for multiple roles, including designers, design directors, merchandisers, planners, and suppliers.
To achieve this, I conducted user research through interviews and shadowing sessions, ensuring I captured the nuances of each role’s workflow. My goal was to anchor the interface in the existing process, using familiar terminology and frameworks, while seamlessly integrating AI enhancements to elevate their capabilities.Throughout the design process, I emphasized a human-centric approach, giving users control over options and editing to maintain a sense of ownership and creativity.
I utilized iterative design techniques, creating wireframes and prototypes in Figma. By incorporating interactive elements and intuitive navigation, I ensured the platform was not only functional but also engaging. Collaborating closely with ML6, our AI development partner, allowed us to align the design within the technical functionalities. This project is currently in progress, and the next phase will involve usability testing to further refine and optimize the interface based on user feedback.
Together, we created an AI-powered engine that outperformed current systems in speed, accuracy, and automation. We focused on producing feasible designs within system constraints, using AI agents to generate garments achievable within desired time frames and aligned with brand identity. To enhance the engine's effectiveness, a synthetic personas feedback interface is under development. Crafting persona cards representing key user segments, especially Gen Z consumers, allows the AI to tailor designs to user preferences.
The synthetic feedback process is designed to mimic how fashion designers currently get feedback, incorporating elements of 1:1 interviews and social listening.Implementing synthetic feedback is an ongoing process, with real-time mechanisms enabling adjustments and prioritizing designs based on purchase intent and trend alignment. This evolving approach ensures the AI engine continues to produce on-trend collections and optimize them for market success.
In this case study, I focused on the product design aspects most relevant to my portfolio. Please note that this project is still ongoing. For more information on the strategy and inner workings of the engine, you can visit BOI's website.
