Powering E-Commerce with a Recommender System Enriched by User’s Emotional Context
Personalization in fashion has been primarily done on segments of consumers using their demographic, geographic and behavioral data, but consumers buy clothes based on their emotions & self-perceptions about their body. In this session, I will highlight how we leverage deep learning and machine learning to understand and decode a consumer’s body perceptions and then personalize the retailer’s digital experience throughout the consumer's lifecycle. In particular, I will share the challenges in gathering training data, ever growing custom models, understanding cognitive attributes of users as we built our AI platform.
Sowmiya is the co-founder/CTO of Lily AI, an emotional intelligence powered shopping experience that helps consumers discover clothes that make them look & feel their best. At Lily, she is focussed on decoding user behavior and building deep product understanding by applying machine learning/deep learning techniques. Prior to Lily, she worked at different levels of the tech stack at Box leading initiatives in - building SDKs, applications for industry verticals & MDM solutions. She was also an early engineer at Pocket Gems where she worked on the core game engine and built acquisition and retention strategies for #1 & #4 top grossing gaming apps. Sowmiya is a UT Austin grad with a Masters in Electrical and Computer Engineering.