From Search to Insight: Leveraging OpenSearch for Scalable, AI-Driven Search Experiences
summary
OpenSearch 3.0 delivers concrete changes: Lucene 10 and JVM 21 yield about 20% faster queries than 2.x and 10× throughput versus 1.x, while ANN/HNSW vector search and GPU-accelerated vector indexing cut index build time (9.3× faster in a 10M-vector benchmark) and lower cost (~3.75×). Derived sources reduce stored vector size and memory pressure, reciprocal rank fusion and neural sparse search improve hybrid ranking, and Arrow/Flight streaming plus concurrent segment search boost distributed query throughput.
AI-written summary · about
chapters (12)
- 0:00Introduction and agenda
- 1:44Evolution of search technology
- 4:36OpenSearch foundation overview
- 8:27OpenSearch 3.0 technical deep dive
- 10:56OpenSearch vector engine architecture
- 15:09Embedding creation with ML connectors
- 17:39Retrieving and ranking embeddings efficiently
- 19:24RAG integration and real-time updates
- 20:52Performance and scalability improvements in 3.0
- 24:22Streaming and cloud-native architecture
- 27:32Future roadmap and community engagement
- 35:18Q&A session
related talks
55mVector Search Made Simple: Getting Started with OpenSearch for AI Applications
46mDelivering the next wave of innovation with OpenSearch
22mVector Search Made Simple: Getting Started With OpenSearch for AI Applications
28mUse OpenSearch as Your Vector Search Database To Optimize GenAI Accuracy!
55mSupercharged Search with Semantic Search and Vector Embeddings
40mAgentic Relevance Tuning With OpenSearch: Faster, Easier, and Higher Quality Search at Scale
38mAI-Ready Data Foundation Using Amazon OpenSearch
36mBuilding Next-gen Search With OpenSearch