Case Study: HerMin Textile achieves 25% faster time-to-market and 10,000+ design output with Google Cloud Platform

HerMin Textile Competing in the fashion market using Google and TensorFlow

HerMin Textile, founded in 1976 in Taiwan, supplies woven fabrics to global apparel brands such as Ralph Lauren, J. Crew, Filson, and Burberry. Facing growing pressure from fast fashion and new designers, the company’s manual process of storing thousands of sample designs in a warehouse and photographing them for customers was slow and inefficient, hindering design sharing and time to market.

With CloudMile’s help, HerMin built a TensorFlow-based machine learning pattern-recognition system and a mobile app on Google Cloud Platform—using Cloud Storage, Cloud Datalab, and Compute Engine—to automate image analysis and modernize designer collaboration. The solution lets the company share and generate similar patterns at scale, enabling more than 10,000 new designs in two years, cutting time to market by about 25% (from ~12 to ~9 months), and paving the way for a SaaS platform and a designer community.

HerMin Textile

Neil Lee

Consultant