Fashion start-up Lalaland will run its e-commerce fashion platform on Amazon Web Services. Virtual fashion models are generated on this platform using AI and machine learning, representing different body types, sizes, hair types and skin tones that are closest to the shopper’s preference.
With ML on AWS, retail, e-commerce and marketing companies can generate, dress and style their own models within 72 hours. This is 230% faster than taking photos and saves up to 70% on typical photography production costs. Lalaland claims that it will allow retailers to increase their sales by 15% thanks to better first-time correct hit rates and the reduction of the need to order different sizes as the virtual fitting experience is better tuned.
To develop their virtual models, Lalaland uses Core ML and Supplementary ML on AWS, specifically the AWS services Sagemaker, DynamoDB and ECS. They combine ML and AI engineering with image recognition and deep learning to provide fashion brands, retailers and shoppers with a diverse library of AI-generated human models.