Keynote: Advancing Search and Observability at Uber with OpenSearch - Shanshan Song, Senior Director of Engineering, Uber Search is a foundational pillar of Uber’s platform, powering critical user experiences across products like Uber Eats, where users navigate vast catalogs—over one million restaurants globally and thousands of options per session. To meet the demands of this complex choice problem, Uber has invested heavily in the evolution of its search infrastructure. A key area of innovation is semantic search, enabling Uber’s systems to better understand user intent and retrieve relevant results beyond keyword matching. Uber engineers have extended OpenSearch to support semantic vector search at scale, integrating real-time retrieval pipelines with relevance-aware ranking models. These enhancements allow Uber to deliver more intuitive and personalized search results across diverse use cases. In addition to search, Uber is also addressing a major gap in observability: metrics. While OpenSearch is widely used for logs and traces, native metric support has been limited. Uber is leading the development of a high-performance time-series engine within OpenSearch, introducing new data structures and compression algorithms optimized for time-based workloads. This engine will support standards like Prometheus and offer pluggable integration with systems like M3. Uber is committed to contributing this work, and innovating with the ecosystem, and being part of this community to advance unified observability solutions.