DeveloperWeek 2019

With more products flooding the market than ever before, it’s imperative that businesses from retail to real estate not only get their own products in front of customers but also ensure they’re delivering the right ones. Recommender systems meet this need by directing customers to desired products and information, providing major benefits to both companies and users.

Unfortunately, the value of these recommender systems is limited by how much (or how little) they understand individual users and how they collect data on their preferences. Without specific and individualized data, companies can’t deliver the information their customers are looking for.
Advances in deep learning and visual search are enabling companies to collect and use data in smarter ways to deliver recommendations that are more personalized and more useful. AI tools use a person’s digital behavior to individualize recommender results, while visual search technology allows users to upload their own photos to further curate recommendations. Trulia is exploring these capabilities to separate user behavior from user intent and to offer new ways for customers to share personal preferences.

The retail and interior design industries are already leveraging these tools to provide more useful recommendations. The potential of this technology spans sectors from real estate to automotive sales to beauty products and beyond.

In this session, Shourabh Rawat will outline the use cases for deep learning and visual search. He will provide best practices for building systems that can leverage these technologies. Audience members will walk away with a clear understanding of the values of these tools and how to develop applications to implement them at their own companies.