Uber Adopts Amazon OpenSearch for Semantic Search to Better Capture User Intent

"The first step in Uber's adoption of OpenSearch was to evaluate it against their existing Lucene-based setup using th HNSW (Hierarchical Navigable Small World) algorithm: We found ourselves limited by the lack of algorithm options, which hindered our ability to fine-tune trade-offs for different scenarios."

"Another key limitation of their existing architecture was the lack of native GPU support, which became a performance bottleneck for personalized recommendations and fraud detection, two complex workloads relying on high-dimensional vectors. Amazon OpenSearch addressed both limitations by supporting multiple Approximate Nearest Neighbor (ANN) algorithms, providing flexibility for different use cases, and offering GPU acceleration through Facebook AI Similarity Search."

"After selecting the technology, Uber built a prototype capable of handling over 1.5 billion vectors across nearly 400 dimensions to enable large-scale semantic retrieval. The prototype processed raw data to generate embeddings, which were stored using Apache Hive. The data was then batch ingested into OpenSearch using Spark and queried with FAISS. The most significant challenges Uber engineers had to overcome were related to ingestion speed and stability, and query performance."

Uber migrated from Apache Lucene to Amazon OpenSearch to support large-scale vector search and better capture search intent. OpenSearch provided multiple ANN algorithms and GPU acceleration via Facebook AI Similarity Search, addressing limitations of algorithm availability and lack of native GPU support that constrained personalization and fraud-detection workloads. A prototype indexed over 1.5 billion vectors across nearly 400 dimensions. The prototype pipeline generated embeddings, stored data in Apache Hive, batch ingested into OpenSearch using Spark, and served queries with FAISS. Engineers prioritized improvements to ingestion speed, stability, and query performance through targeted infrastructure optimizations.

Read at InfoQ