Imagine a future where fashion designers are no longer solitary artists but "style prophets" equipped with AI assistants that can accurately predict trends, efficiently design hit products, and even intelligently manage inventory to ensure every garment finds its perfect home. This isn't science fiction—it's the new vision for the apparel industry being realized by Zhiyi Technology CEO Zheng Zeyu.
In 2015, Zheng co-founded Caicloud with several Google colleagues, focusing on providing AI solutions for enterprises. While collaborating with Chinese MCN agencies like Ruhan, he identified profound changes occurring in the fashion sector. Data analysis revealed that between 2016-2017, consumer purchasing decisions underwent a significant shift—buyers stopped blindly following traditional brands and began placing greater trust in recommendations from KOLs and KOCs. This power restructuring within the industry presented substantial business opportunities. Zheng also noted that despite established players like Uniqlo holding significant market share, the apparel industry lacked true monopolists, indicating enormous growth potential. Based on these insights, Zheng founded Zhiyi Technology in 2018 to explore platform-based opportunities in fashion.
AI's Transformative Impact on Fashion
According to Statista, the global fashion apparel market is projected to reach $770.9 billion in 2024, growing at a compound annual rate of 8.94% through 2029. As large language models and generative AI disrupt industries at unprecedented speed, fashion stands poised for transformation.
"Our company name reflects our core philosophy—using technology to drive fashion innovation," explains Zheng. "Since 2018, I've constantly been asked how AI can assist the apparel industry. Many designers view fashion as pure artistic creation rather than technology-driven, which creates skepticism about AI applications."
Zheng's background includes an AI master's degree from Carnegie Mellon University and experience at Microsoft Research and Google. His first startup provided AI outsourcing services across industries before specializing in fashion. "We realized clothing design isn't just art—it's fundamentally merchandise. As products, designs must consider market demand and consumer preferences, making them perfect for data analysis and market research."
Generative AI's New Possibilities
The advent of large language models enables capabilities previously unimaginable:
- Smarter Tagging Systems: Traditional methods struggled to extract key information from innovative designs. AI models now rapidly construct flexible tagging systems that extract and refine elements from unstructured data, identifying trends and insights.
- Designer Empowerment: Where creatives once relied solely on imagination, AIGC technology now enables automatic combinations of design elements. More crucially, AI analyzes designers' behaviors, historical data, positioning and personal styles to generate elements aligned with current trends—serving as a 24/7 "muse" for inspiration.
Solving Industry Challenges
Women's fashion faces particular complexity—numerous long-tail brands, rampant design copying, and high return rates. AI currently addresses two key issues:
- Precision Design: Brands often accumulate excess inventory from miscalculated designs. AI tools provide market analysis to help designers avoid missteps—functioning as a "market radar" for target audience alignment.
- Supply Chain Coordination: The platform identifies potentially popular products, enabling small test batches before scaling production via quick-response manufacturing. Zhiyi has connected over 1,000 supply chain partners including fabric suppliers and garment factories to support this agile system.
High return rates present more complex challenges involving multiple factors beyond any single technological solution. Current focus remains on ensuring designed products sell successfully while optimizing inventory through supply chain coordination.
Integration and Implementation
Zhiyi operates a complete retail management system and requires ecosystem partners to adopt its platform. "The apparel industry's fragmentation means many factories lack existing systems or resist using them," Zheng notes. "Like Shein—which manages $30 billion+ GMV with just 2,000 employees—we aim to build similarly efficient supply chain management."
Key integration challenges include consolidating orders across dispersed supply chains and ensuring data accuracy. "Some suppliers might falsely report completed deliveries. Ensuring truthful, timely data requires solutions across system architecture and process management—an ongoing optimization process."
Industry Reception and Future Vision
While most fashion companies show strong interest in AI design applications, designers express both excitement and apprehension about potential job displacement. Technical hurdles remain in achieving precise, controllable outputs that match designers' styles and expectations.
For smaller brands with limited data, Zhiyi's solution aggregates industry-wide information rather than relying on individual company data. "Our database contains hundreds of billions of apparel-related data points, enabling personalized style generation based on each client's positioning," Zheng explains.
Contrary to concerns about increased design copying, AI may actually reduce homogeneity. "Most designers don't possess extraordinary creativity—they traditionally spend hours tracking competitors. With AI providing access to thousands of data sources instead of just 10-20 brands, designs become more distinctive." As generative AI expands information access from thousands to potentially millions of sources, industry differentiation should increase.
Looking ahead, Zheng envisions AI creating new market opportunities by enhancing brand differentiation. "Frequent product updates with accurate demand prediction can reduce inventory while increasing customer satisfaction and loyalty." The technology also enables virtual try-ons, design feedback, automated copywriting, and varied product displays to meet diverse client preferences.
Future developments could include chatbot interfaces where designers simply converse with AI to understand trends and user needs. "This would democratize access, giving all designers—regardless of company size—equal competitive platforms," Zheng suggests.
The ultimate goal involves digitizing the entire supply chain. "Currently, design, production and sales data exist in silos. If we can integrate this information while protecting sensitive data—what we call 'conservative openness'—we'll achieve true digital transformation over the next 3-5 years."
Defining Zhiyi's Mission
"We aim to build a complete industry ecosystem where tools serve as the core enabler," Zheng states. "Whether through AI, big data, cloud computing or future technologies, our ultimate objective remains enhancing overall industry efficiency and profitability for every supply chain participant."
Recognizing this as a long-term endeavor potentially spanning decades, Zheng sees unique timing: "The industry is transforming. Traditional consumer perceptions about premium brands are breaking down while new fashion paradigms haven't fully emerged. As consumers prioritize individuality, trends are decentralizing. The question becomes: how do we redefine fashion's benchmarks and premium positioning in this new era?"
With economic recovery and global changes ahead, Zhiyi positions itself to help shape fashion's next chapter—not through traditional top-down models, but by empowering designers and consumers alike in an increasingly democratized industry landscape.