A Large-Scale Deep Architecture for Personalized Grocery Basket Recommendations

TL;DR

This paper presents RTT2Vec, a large-scale deep learning system for real-time personalized grocery recommendations that significantly improves prediction accuracy and inference speed, enhancing user experience in online grocery shopping.

Contribution

Introduces a novel deep architecture for real-time personalized grocery recommendations with an efficient inference method, demonstrating substantial performance improvements.

Findings

9.4% uplift in prediction metrics over baseline models

11.6x faster approximate inference technique

Increased basket size and improved product discovery in production

Abstract

With growing consumer adoption of online grocery shopping through platforms such as Amazon Fresh, Instacart, and Walmart Grocery, there is a pressing business need to provide relevant recommendations throughout the customer journey. In this paper, we introduce a production within-basket grocery recommendation system, RTT2Vec, which generates real-time personalized product recommendations to supplement the user's current grocery basket. We conduct extensive offline evaluation of our system and demonstrate a 9.4% uplift in prediction metrics over baseline state-of-the-art within-basket recommendation models. We also propose an approximate inference technique 11.6x times faster than exact inference approaches. In production, our system has resulted in an increase in average basket size, improved product discovery, and enabled faster user check-out