Scaling Multilingual Semantic Search In Uber Eats Delivery

Abstract

We present a production-oriented semantic retrieval system for Uber Eats that unifies retrieval across stores, dishes, and grocery/retail items. Our approach fine-tunes a Qwen2 two-tower base model using hundreds of millions of query-document interactions that were aggregated and anonymized pretraining. We train the model with a combination of InfoNCE on in-batch negatives and triplet-NCE loss on hard negatives, and we leverage Matryoshka Representation Learning (MRL) to serve multiple embedding sizes from a single model. Our system achieves substantial recall gains over a strong baseline across six markets and three verticals. This paper presents the end to end work including data curation, model architecture, large-scale training, and evaluation. We also share key insights and practical lessons for building a unified, multilingual, and multi-vertical retrieval system for consumer search.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyling2026scaling

Related papers

• MRSE: An Efficient Multi-modality Retrieval System For Large Scale E-commerce (2024)0.00
• Unified Embedding Based Personalized Retrieval In Etsy Search (2023)2.26
• Multimodal Generative Retrieval Model With Staged Pretraining For Food Delivery On Meituan (2026)0.00
• Unified Learning-to-rank For Multi-channel Retrieval In Large-scale E-commerce Search (2026)0.00
• Zero-shot Retrieval For Scalable Visual Search In A Two-sided Marketplace (2025)1.57
• Multimodal Semantic Retrieval For Product Search (2025)3.58
• Mine And Refine: Optimizing Graded Relevance In E-commerce Search Retrieval (2026)0.00
• Unifier: A Unified Retriever For Large-scale Retrieval (2022)7.50