- Hangzhou startup showcases AI and 3D tools aimed at reshaping apparel design and manufacturing workflows
- It also introduces a “digital co-worker” tool as it expands from fashion software into broader AI applications
AI and 3D simulation will reshape how garments are designed and produced, shifting the industry from forecast-driven manufacturing to demand-led production, an entrepreneur said at a conference in Hangzhou on June 4.
“AI perceives demand, 3D digital twins validate it, and intelligent manufacturing brings it into production. Together, they will drive the industry from forecast-based inventory business to real-time production based on confirmed demand,” Eric Liu Chen, founder and chief executive of Style3D, a Hangzhou-based fashion tech startup, told the audience at the Third Style3D Partner Conference.
The company is positioning its Style3D platform as a digital infrastructure layer for apparel supply chains, spanning design, development, production and cross-company collaboration, Liu said.
Pain points
Traditional apparel development typically follows a linear process from design sketches to 2D pattern-making, physical sampling and repeated revisions.
A single sample can take up to 72 hours to produce, cost around 1,500 yuan, and often requires multiple rounds of international shipping for confirmation, creating delays and inefficiencies across the supply chain.
Style3D is built around a proprietary soft-body simulation engine that converts 2D sketches into production-ready 3D garments in real time.
The system simulates fabric physics and wearing effects, enabling virtual pattern-making, online collaboration and cloud-based design review.
The company said this approach can reduce sample development cycles by more than 80% and cut design costs by up to 90%, compressing the workflow between concept and production.
StyleClaw, a digital co-worker?
During the conference, Style3D officially launched StyleClaw. It is a digital partner platform inspired by the OpenClaw concept, designed for professional apparel workflows.
The system targets tasks including design styles, fabric selection, craftsmanship, technical packages, customer communication and organizational knowledge management.
The platform is structured around “one entry point and multiple expert agents,” which the company says enables enterprises to run a 24/7 digital workforce of AI collaborators with expanding capabilities.
The launch of StyleClaw signals a shift in how AI is positioned within the apparel industry, from a passive tool to an active collaborative agent, or even a co-worker.
Efficiency gains, enhanced conversion
In a case study presented at the event, Chinese menswear brand Baoxiniao showcased its AI agent-driven customization system, which integrates intelligent diagnostics, design generation, virtual try-on and recommendation tools.
“The system improved consumer decision efficiency and increased order conversion rates by 25%,” Zhao Guohua, head of quality supervision at the Wenzhou-based apparel maker.
Wangsida, a garment manufacturer in Quanzhou, Fujian Province, said it cut its production cycle from design to delivery from 15 days to 7 days after adopting AI-enabled 3D design tools.
This approach has also delivered an 80% improvement, following end-to-end digitization of its workflow, Chen Zhenxing, general manager of Wangsida, said.
Venturing into embodied AI
Founded in 2015, Style3D positions itself as an AI and 3D-driven digitization company for the apparel sector.
It has developed a proprietary soft-body simulation engine that it says breaks long-standing Western dominance in garment 3D design software and supports full-chain solutions from design to marketing.
The company serves more than 3,000 clients, including Anta, Bosideng and Nike. It counts among its backers top-tier investors such as Hillhouse Capital, Shunwei Capital, Vision Plus Capital and CDH Investments.
Drawing on years of accumulated data and expertise in simulating deformable materials such as textiles, Style3D has recently branched out into embodied AI.
With the launch of its SynReal platform, it is diversifying its product portfolio into a simulation software stack for data collection and training on deformable objects.
